2019-10-09 00:34:25.536: INFO @api : Creating dataset 2019-10-09 00:34:27.503: INFO @api : RecognitionDatasetWrapper created 2019-10-09 00:34:27.503: INFO @api : Creating a model 2019-10-09 00:34:27.523: WARNING @deprecation_wrapper: From /home/red/emloop-tensorflow/emloop_tensorflow/utils/profiler.py:12: The name tf.Session is deprecated. Please use tf.compat.v1.Session instead. 2019-10-09 00:34:27.860: WARNING @model : Quantization trainign is experimental atm, please read the warning in the docs first. 2019-10-09 00:34:27.861: INFO @model : Creating TF model on 1 GPU devices 2019-10-09 00:34:27.861: WARNING @deprecation_wrapper: From /home/red/emloop-tensorflow/emloop_tensorflow/model.py:415: The name tf.ConfigProto is deprecated. Please use tf.compat.v1.ConfigProto instead. 2019-10-09 00:34:28.299: WARNING @deprecation_wrapper: From /home/red/emloop-tensorflow/emloop_tensorflow/model.py:140: The name tf.variable_scope is deprecated. Please use tf.compat.v1.variable_scope instead. 2019-10-09 00:34:28.299: WARNING @deprecation_wrapper: From /home/red/emloop-tensorflow/emloop_tensorflow/model.py:140: The name tf.get_variable_scope is deprecated. Please use tf.compat.v1.get_variable_scope instead. 2019-10-09 00:34:28.299: WARNING @deprecation_wrapper: From /home/red/emloop-tensorflow/emloop_tensorflow/model.py:140: The name tf.AUTO_REUSE is deprecated. Please use tf.compat.v1.AUTO_REUSE instead. 2019-10-09 00:34:29.263: WARNING @deprecation : From /home/red/faces/models/recognition.py:89: flatten (from tensorflow.python.layers.core) is deprecated and will be removed in a future version. Instructions for updating: Use keras.layers.flatten instead. 2019-10-09 00:34:29.428: WARNING @deprecation : From /home/red/faces/models/recognition.py:90: dropout (from tensorflow.python.layers.core) is deprecated and will be removed in a future version. Instructions for updating: Use keras.layers.dropout instead. 2019-10-09 00:34:29.474: WARNING @deprecation : From /home/red/faces/models/recognition.py:91: dense (from tensorflow.python.layers.core) is deprecated and will be removed in a future version. Instructions for updating: Use keras.layers.dense instead. 2019-10-09 00:34:29.476: WARNING @deprecation : From /usr/lib/python3.7/site-packages/tensorflow/python/ops/init_ops.py:1251: calling VarianceScaling.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version. Instructions for updating: Call initializer instance with the dtype argument instead of passing it to the constructor 2019-10-09 00:34:29.753: WARNING @deprecation : From /usr/lib/python3.7/site-packages/tensorflow/python/ops/array_ops.py:1354: add_dispatch_support..wrapper (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version. Instructions for updating: Use tf.where in 2.0, which has the same broadcast rule as np.where 2019-10-09 00:34:29.885: INFO @model : Applying quantization aware training updates with 0 delay 2019-10-09 00:34:30.688: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:34:30.717: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:34:30.746: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:34:30.775: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:34:30.804: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:34:30.833: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:34:30.863: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:34:30.891: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:34:30.922: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:34:30.951: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:34:30.980: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:34:31.010: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:34:31.040: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:34:31.069: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:34:31.098: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:34:31.130: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:34:31.159: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:34:31.189: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:34:31.218: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:34:31.248: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:34:31.278: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:34:31.308: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:34:31.337: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:34:31.367: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:34:31.397: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:34:31.427: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:34:31.565: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 00:34:33.381: WARNING @deprecation_wrapper: From /home/red/emloop-tensorflow/emloop_tensorflow/model.py:180: The name tf.train.Saver is deprecated. Please use tf.compat.v1.train.Saver instead. 2019-10-09 00:34:34.730: INFO @model : Number of parameters: 1042605 2019-10-09 00:34:34.731: INFO @api : RecognitionModel created 2019-10-09 00:34:34.731: INFO @api : Creating hooks 2019-10-09 00:34:34.738: INFO @api : ComputeStats created 2019-10-09 00:34:34.739: INFO @api : LogVariables created 2019-10-09 00:34:34.739: INFO @api : LogProfile created 2019-10-09 00:34:34.740: INFO @api : LogDir created 2019-10-09 00:34:34.748: INFO @api : ComputeMatchingMetrics created 2019-10-09 00:34:34.748: INFO @api : ComputeRecognitionMetrics created 2019-10-09 00:34:34.748: WARNING @deprecation_wrapper: From /home/red/emloop-tensorflow/emloop_tensorflow/hooks/write_tensorboard.py:78: The name tf.summary.FileWriter is deprecated. Please use tf.compat.v1.summary.FileWriter instead. 2019-10-09 00:34:34.748: INFO @api : WriteTensorBoard created 2019-10-09 00:34:34.749: INFO @api : DecayLR created 2019-10-09 00:34:34.749: INFO @api : SaveBest created 2019-10-09 00:34:34.750: INFO @api : StopAfter created 2019-10-09 00:34:34.751: INFO @api : ShowProgress created 2019-10-09 00:34:34.751: WARNING @api : TrainingTrace hook added between hooks. Add it to your config.yaml to suppress this warning. 2019-10-09 00:34:34.751: INFO @api : Creating main loop 2019-10-09 00:34:34.751: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 00:34:34.752: INFO @main_loop : Training epoch 1 2019-10-09 00:34:38.234: WARNING @main_loop : Some sources provided by the stream do not match model placeholders. Set `main_loop.on_unused_sources` to `ignore` in order to suppress this warning. Extra sources: ['box_points', 'filenames'] 2019-10-09 00:38:02.141: INFO @log_variables: train loss mean: 0.984897 2019-10-09 00:38:02.141: INFO @log_variables: train age_loss mean: 14.955156 2019-10-09 00:38:02.141: INFO @log_variables: train gender_loss mean: 0.661527 2019-10-09 00:38:02.141: INFO @log_variables: train matching_loss nanmean: 0.875633 2019-10-09 00:38:02.141: INFO @log_variables: train is_face_loss mean: 0.020507 2019-10-09 00:38:02.141: INFO @log_variables: train age_mae mean: 15.447821 2019-10-09 00:38:02.141: INFO @log_variables: train gender_accuracy mean: 0.624795 2019-10-09 00:38:02.141: INFO @log_variables: train positive_distance nanmean: 0.785072 2019-10-09 00:38:02.141: INFO @log_variables: train negative_distance nanmean: 1.396594 2019-10-09 00:38:02.141: INFO @log_variables: train is_face_accuracy mean: 0.992952 2019-10-09 00:38:02.141: INFO @log_variables: valid loss mean: 0.728190 2019-10-09 00:38:02.141: INFO @log_variables: valid age_loss mean: 9.280687 2019-10-09 00:38:02.141: INFO @log_variables: valid gender_loss mean: 0.533090 2019-10-09 00:38:02.141: INFO @log_variables: valid matching_loss nanmean: 0.796228 2019-10-09 00:38:02.141: INFO @log_variables: valid is_face_loss mean: 0.000002 2019-10-09 00:38:02.142: INFO @log_variables: valid age_mae mean: 9.768431 2019-10-09 00:38:02.142: INFO @log_variables: valid gender_accuracy mean: 0.716880 2019-10-09 00:38:02.142: INFO @log_variables: valid positive_distance nanmean: 0.703471 2019-10-09 00:38:02.142: INFO @log_variables: valid negative_distance nanmean: 1.197024 2019-10-09 00:38:02.142: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 00:38:02.142: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 00:38:03.622: INFO @metrics_hook: valid matching accuracy: 0.7626071835462014, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 00:38:03.998: INFO @decay_lr : LR updated to `0.000995` 2019-10-09 00:38:03.999: WARNING @deprecation_wrapper: From /home/red/emloop-tensorflow/emloop_tensorflow/model.py:301: The name tf.train.write_graph is deprecated. Please use tf.io.write_graph instead. 2019-10-09 00:38:04.848: INFO @model : Quantizing and saving the model 2019-10-09 00:38:05.743: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:38:05.748: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:38:05.753: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:38:05.758: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:38:05.763: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:38:05.768: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:38:05.773: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:38:05.778: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:38:05.783: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:38:05.788: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:38:05.794: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:38:05.799: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:38:05.804: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:38:05.809: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:38:05.814: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:38:05.819: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:38:05.824: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:38:05.829: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:38:05.834: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:38:05.839: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:38:05.844: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:38:05.849: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:38:05.854: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:38:05.859: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:38:05.864: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:38:05.869: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:38:05.874: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 00:38:05.881: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 00:38:30.861: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 00:38:30.991: WARNING @deprecation : From /usr/lib/python3.7/site-packages/tensorflow/lite/python/util.py:238: convert_variables_to_constants (from tensorflow.python.framework.graph_util_impl) is deprecated and will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.convert_variables_to_constants` 2019-10-09 00:38:30.991: WARNING @deprecation : From /usr/lib/python3.7/site-packages/tensorflow/python/framework/graph_util_impl.py:270: extract_sub_graph (from tensorflow.python.framework.graph_util_impl) is deprecated and will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.extract_sub_graph` 2019-10-09 00:38:31.100: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 00:38:31.118: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 00:38:32.920: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 00:38:32.963: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 00:38:32.967: INFO @log_profile : T train: 192.659206 2019-10-09 00:38:32.968: INFO @log_profile : T valid: 9.768977 2019-10-09 00:38:32.968: INFO @log_profile : T read data: 4.392845 2019-10-09 00:38:32.969: INFO @log_profile : T hooks: 31.311822 2019-10-09 00:38:32.969: INFO @main_loop : Epoch 1 done 2019-10-09 00:38:32.969: INFO @main_loop : Training epoch 2 2019-10-09 00:41:51.079: INFO @log_variables: train loss mean: 0.739601 2019-10-09 00:41:51.079: INFO @log_variables: train age_loss mean: 10.635652 2019-10-09 00:41:51.079: INFO @log_variables: train gender_loss mean: 0.473146 2019-10-09 00:41:51.079: INFO @log_variables: train matching_loss nanmean: 0.755993 2019-10-09 00:41:51.079: INFO @log_variables: train is_face_loss mean: 0.000060 2019-10-09 00:41:51.079: INFO @log_variables: train age_mae mean: 11.125751 2019-10-09 00:41:51.079: INFO @log_variables: train gender_accuracy mean: 0.766299 2019-10-09 00:41:51.079: INFO @log_variables: train positive_distance nanmean: 0.834623 2019-10-09 00:41:51.079: INFO @log_variables: train negative_distance nanmean: 1.407983 2019-10-09 00:41:51.079: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 00:41:51.079: INFO @log_variables: valid loss mean: 0.649199 2019-10-09 00:41:51.079: INFO @log_variables: valid age_loss mean: 8.913258 2019-10-09 00:41:51.079: INFO @log_variables: valid gender_loss mean: 0.402073 2019-10-09 00:41:51.079: INFO @log_variables: valid matching_loss nanmean: 0.719113 2019-10-09 00:41:51.080: INFO @log_variables: valid is_face_loss mean: 0.000005 2019-10-09 00:41:51.080: INFO @log_variables: valid age_mae mean: 9.402090 2019-10-09 00:41:51.080: INFO @log_variables: valid gender_accuracy mean: 0.808375 2019-10-09 00:41:51.080: INFO @log_variables: valid positive_distance nanmean: 0.746947 2019-10-09 00:41:51.080: INFO @log_variables: valid negative_distance nanmean: 1.265531 2019-10-09 00:41:51.080: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 00:41:51.080: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 00:41:52.980: INFO @metrics_hook: valid matching accuracy: 0.8106973676320681, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 00:41:53.325: INFO @decay_lr : LR updated to `0.000990025` 2019-10-09 00:41:53.875: INFO @model : Quantizing and saving the model 2019-10-09 00:41:54.651: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:41:54.656: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:41:54.661: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:41:54.666: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:41:54.671: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:41:54.675: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:41:54.680: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:41:54.685: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:41:54.690: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:41:54.695: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:41:54.700: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:41:54.705: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:41:54.710: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:41:54.715: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:41:54.720: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:41:54.725: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:41:54.730: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:41:54.735: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:41:54.740: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:41:54.745: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:41:54.750: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:41:54.755: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:41:54.760: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:41:54.765: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:41:54.770: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:41:54.775: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:41:54.780: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 00:41:54.787: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 00:42:04.785: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 00:42:05.020: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 00:42:05.039: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 00:42:06.924: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 00:42:06.959: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 00:42:06.963: INFO @log_profile : T train: 187.123826 2019-10-09 00:42:06.963: INFO @log_profile : T valid: 8.950299 2019-10-09 00:42:06.964: INFO @log_profile : T read data: 1.366725 2019-10-09 00:42:06.964: INFO @log_profile : T hooks: 16.469178 2019-10-09 00:42:06.964: INFO @main_loop : Epoch 2 done 2019-10-09 00:42:06.964: INFO @main_loop : Training epoch 3 2019-10-09 00:45:24.673: INFO @log_variables: train loss mean: 0.664511 2019-10-09 00:45:24.674: INFO @log_variables: train age_loss mean: 9.874996 2019-10-09 00:45:24.674: INFO @log_variables: train gender_loss mean: 0.384250 2019-10-09 00:45:24.674: INFO @log_variables: train matching_loss nanmean: 0.688174 2019-10-09 00:45:24.674: INFO @log_variables: train is_face_loss mean: 0.000062 2019-10-09 00:45:24.674: INFO @log_variables: train age_mae mean: 10.364021 2019-10-09 00:45:24.674: INFO @log_variables: train gender_accuracy mean: 0.822938 2019-10-09 00:45:24.674: INFO @log_variables: train positive_distance nanmean: 0.850053 2019-10-09 00:45:24.674: INFO @log_variables: train negative_distance nanmean: 1.410637 2019-10-09 00:45:24.674: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 00:45:24.674: INFO @log_variables: valid loss mean: 0.595293 2019-10-09 00:45:24.674: INFO @log_variables: valid age_loss mean: 8.132525 2019-10-09 00:45:24.674: INFO @log_variables: valid gender_loss mean: 0.350088 2019-10-09 00:45:24.674: INFO @log_variables: valid matching_loss nanmean: 0.682065 2019-10-09 00:45:24.674: INFO @log_variables: valid is_face_loss mean: 0.000004 2019-10-09 00:45:24.674: INFO @log_variables: valid age_mae mean: 8.619793 2019-10-09 00:45:24.674: INFO @log_variables: valid gender_accuracy mean: 0.838952 2019-10-09 00:45:24.674: INFO @log_variables: valid positive_distance nanmean: 0.760590 2019-10-09 00:45:24.675: INFO @log_variables: valid negative_distance nanmean: 1.286051 2019-10-09 00:45:24.675: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 00:45:24.675: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 00:45:27.133: INFO @metrics_hook: valid matching accuracy: 0.8327037236913114, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 00:45:27.521: INFO @decay_lr : LR updated to `0.0009850749` 2019-10-09 00:45:28.194: INFO @model : Quantizing and saving the model 2019-10-09 00:45:29.290: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:45:29.296: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:45:29.302: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:45:29.309: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:45:29.314: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:45:29.320: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:45:29.325: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:45:29.331: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:45:29.337: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:45:29.342: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:45:29.348: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:45:29.353: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:45:29.359: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:45:29.364: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:45:29.370: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:45:29.376: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:45:29.382: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:45:29.388: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:45:29.393: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:45:29.398: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:45:29.404: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:45:29.410: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:45:29.416: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:45:29.421: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:45:29.427: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:45:29.433: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:45:29.438: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 00:45:29.446: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 00:45:41.729: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 00:45:42.027: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 00:45:42.047: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 00:45:43.966: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 00:45:44.012: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 00:45:44.016: INFO @log_profile : T train: 187.193931 2019-10-09 00:45:44.016: INFO @log_profile : T valid: 8.862298 2019-10-09 00:45:44.017: INFO @log_profile : T read data: 1.009542 2019-10-09 00:45:44.017: INFO @log_profile : T hooks: 19.903383 2019-10-09 00:45:44.017: INFO @main_loop : Epoch 3 done 2019-10-09 00:45:44.017: INFO @main_loop : Training epoch 4 2019-10-09 00:49:02.296: INFO @log_variables: train loss mean: 0.615160 2019-10-09 00:49:02.297: INFO @log_variables: train age_loss mean: 9.336602 2019-10-09 00:49:02.297: INFO @log_variables: train gender_loss mean: 0.332667 2019-10-09 00:49:02.297: INFO @log_variables: train matching_loss nanmean: 0.640623 2019-10-09 00:49:02.297: INFO @log_variables: train is_face_loss mean: 0.000047 2019-10-09 00:49:02.297: INFO @log_variables: train age_mae mean: 9.825105 2019-10-09 00:49:02.297: INFO @log_variables: train gender_accuracy mean: 0.852563 2019-10-09 00:49:02.297: INFO @log_variables: train positive_distance nanmean: 0.847182 2019-10-09 00:49:02.297: INFO @log_variables: train negative_distance nanmean: 1.410895 2019-10-09 00:49:02.297: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 00:49:02.297: INFO @log_variables: valid loss mean: 0.591319 2019-10-09 00:49:02.297: INFO @log_variables: valid age_loss mean: 8.642199 2019-10-09 00:49:02.297: INFO @log_variables: valid gender_loss mean: 0.307180 2019-10-09 00:49:02.297: INFO @log_variables: valid matching_loss nanmean: 0.661682 2019-10-09 00:49:02.297: INFO @log_variables: valid is_face_loss mean: 0.000007 2019-10-09 00:49:02.297: INFO @log_variables: valid age_mae mean: 9.130217 2019-10-09 00:49:02.298: INFO @log_variables: valid gender_accuracy mean: 0.865744 2019-10-09 00:49:02.298: INFO @log_variables: valid positive_distance nanmean: 0.769071 2019-10-09 00:49:02.298: INFO @log_variables: valid negative_distance nanmean: 1.298473 2019-10-09 00:49:02.298: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 00:49:02.298: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 00:49:04.485: INFO @metrics_hook: valid matching accuracy: 0.8391197457576303, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 00:49:04.855: INFO @decay_lr : LR updated to `0.0009801495` 2019-10-09 00:49:05.473: INFO @model : Quantizing and saving the model 2019-10-09 00:49:06.376: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:49:06.382: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:49:06.388: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:49:06.394: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:49:06.400: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:49:06.405: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:49:06.412: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:49:06.417: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:49:06.422: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:49:06.428: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:49:06.434: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:49:06.439: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:49:06.445: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:49:06.450: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:49:06.456: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:49:06.462: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:49:06.468: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:49:06.474: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:49:06.480: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:49:06.485: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:49:06.491: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:49:06.497: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:49:06.502: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:49:06.508: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:49:06.515: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:49:06.521: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:49:06.526: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 00:49:06.534: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 00:49:17.596: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 00:49:17.849: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 00:49:17.868: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 00:49:19.811: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 00:49:19.847: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 00:49:19.850: INFO @log_profile : T train: 187.282036 2019-10-09 00:49:19.850: INFO @log_profile : T valid: 8.893516 2019-10-09 00:49:19.851: INFO @log_profile : T read data: 1.459053 2019-10-09 00:49:19.851: INFO @log_profile : T hooks: 18.114009 2019-10-09 00:49:19.851: INFO @main_loop : Epoch 4 done 2019-10-09 00:49:19.851: INFO @main_loop : Training epoch 5 2019-10-09 00:52:37.440: INFO @log_variables: train loss mean: 0.578708 2019-10-09 00:52:37.440: INFO @log_variables: train age_loss mean: 8.857740 2019-10-09 00:52:37.440: INFO @log_variables: train gender_loss mean: 0.299490 2019-10-09 00:52:37.440: INFO @log_variables: train matching_loss nanmean: 0.608684 2019-10-09 00:52:37.440: INFO @log_variables: train is_face_loss mean: 0.000045 2019-10-09 00:52:37.440: INFO @log_variables: train age_mae mean: 9.345720 2019-10-09 00:52:37.440: INFO @log_variables: train gender_accuracy mean: 0.869877 2019-10-09 00:52:37.440: INFO @log_variables: train positive_distance nanmean: 0.842791 2019-10-09 00:52:37.440: INFO @log_variables: train negative_distance nanmean: 1.411065 2019-10-09 00:52:37.440: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 00:52:37.440: INFO @log_variables: valid loss mean: 0.555931 2019-10-09 00:52:37.440: INFO @log_variables: valid age_loss mean: 7.856136 2019-10-09 00:52:37.440: INFO @log_variables: valid gender_loss mean: 0.322594 2019-10-09 00:52:37.440: INFO @log_variables: valid matching_loss nanmean: 0.615172 2019-10-09 00:52:37.440: INFO @log_variables: valid is_face_loss mean: 0.000006 2019-10-09 00:52:37.440: INFO @log_variables: valid age_mae mean: 8.341685 2019-10-09 00:52:37.441: INFO @log_variables: valid gender_accuracy mean: 0.854211 2019-10-09 00:52:37.441: INFO @log_variables: valid positive_distance nanmean: 0.797838 2019-10-09 00:52:37.441: INFO @log_variables: valid negative_distance nanmean: 1.331644 2019-10-09 00:52:37.441: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 00:52:37.441: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 00:52:39.647: INFO @metrics_hook: valid matching accuracy: 0.8484739461533849, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 00:52:40.023: INFO @decay_lr : LR updated to `0.00097524875` 2019-10-09 00:52:40.024: INFO @log_profile : T train: 186.860255 2019-10-09 00:52:40.024: INFO @log_profile : T valid: 8.675289 2019-10-09 00:52:40.024: INFO @log_profile : T read data: 1.410317 2019-10-09 00:54:34.229: INFO @log_profile : T hooks: 3.141578 2019-10-09 00:54:34.229: INFO @main_loop : Epoch 5 done 2019-10-09 00:54:34.229: INFO @main_loop : Training epoch 6 2019-10-09 00:57:42.601: INFO @log_variables: train loss mean: 0.555875 2019-10-09 00:57:42.601: INFO @log_variables: train age_loss mean: 8.572196 2019-10-09 00:57:42.601: INFO @log_variables: train gender_loss mean: 0.280647 2019-10-09 00:57:42.601: INFO @log_variables: train matching_loss nanmean: 0.585314 2019-10-09 00:57:42.601: INFO @log_variables: train is_face_loss mean: 0.000030 2019-10-09 00:57:42.601: INFO @log_variables: train age_mae mean: 9.059474 2019-10-09 00:57:42.601: INFO @log_variables: train gender_accuracy mean: 0.878913 2019-10-09 00:57:42.601: INFO @log_variables: train positive_distance nanmean: 0.837649 2019-10-09 00:57:42.601: INFO @log_variables: train negative_distance nanmean: 1.411098 2019-10-09 00:57:42.602: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 00:57:42.602: INFO @log_variables: valid loss mean: 0.532094 2019-10-09 00:57:42.602: INFO @log_variables: valid age_loss mean: 7.586518 2019-10-09 00:57:42.602: INFO @log_variables: valid gender_loss mean: 0.273643 2019-10-09 00:57:42.602: INFO @log_variables: valid matching_loss nanmean: 0.617194 2019-10-09 00:57:42.602: INFO @log_variables: valid is_face_loss mean: 0.000003 2019-10-09 00:57:42.602: INFO @log_variables: valid age_mae mean: 8.072604 2019-10-09 00:57:42.602: INFO @log_variables: valid gender_accuracy mean: 0.882482 2019-10-09 00:57:42.602: INFO @log_variables: valid positive_distance nanmean: 0.783175 2019-10-09 00:57:42.602: INFO @log_variables: valid negative_distance nanmean: 1.325370 2019-10-09 00:57:42.602: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 00:57:42.602: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 00:57:44.915: INFO @metrics_hook: valid matching accuracy: 0.851951789890268, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 00:57:45.275: INFO @decay_lr : LR updated to `0.0009703725` 2019-10-09 00:57:45.887: INFO @model : Quantizing and saving the model 2019-10-09 00:57:46.943: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:57:46.949: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:57:46.953: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:57:46.958: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:57:46.963: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:57:46.968: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:57:46.974: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:57:46.979: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:57:46.984: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:57:46.989: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:57:46.994: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:57:47.000: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:57:47.006: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:57:47.011: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:57:47.016: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:57:47.021: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:57:47.026: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:57:47.031: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:57:47.036: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:57:47.041: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:57:47.046: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:57:47.051: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:57:47.057: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:57:47.062: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:57:47.067: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 00:57:47.073: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 00:57:47.078: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 00:57:47.085: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 00:57:58.489: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 00:57:58.756: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 00:57:58.776: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 00:58:00.699: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 00:58:00.733: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 00:58:00.737: INFO @log_profile : T train: 178.460634 2019-10-09 00:58:00.737: INFO @log_profile : T valid: 8.290340 2019-10-09 00:58:00.737: INFO @log_profile : T read data: 0.951082 2019-10-09 00:58:00.737: INFO @log_profile : T hooks: 18.721441 2019-10-09 00:58:00.738: INFO @main_loop : Epoch 6 done 2019-10-09 00:58:00.738: INFO @main_loop : Training epoch 7 2019-10-09 01:01:10.682: INFO @log_variables: train loss mean: 0.535562 2019-10-09 01:01:10.682: INFO @log_variables: train age_loss mean: 8.268322 2019-10-09 01:01:10.683: INFO @log_variables: train gender_loss mean: 0.264769 2019-10-09 01:01:10.683: INFO @log_variables: train matching_loss nanmean: 0.568612 2019-10-09 01:01:10.683: INFO @log_variables: train is_face_loss mean: 0.000028 2019-10-09 01:01:10.683: INFO @log_variables: train age_mae mean: 8.754898 2019-10-09 01:01:10.683: INFO @log_variables: train gender_accuracy mean: 0.887645 2019-10-09 01:01:10.683: INFO @log_variables: train positive_distance nanmean: 0.837832 2019-10-09 01:01:10.683: INFO @log_variables: train negative_distance nanmean: 1.411268 2019-10-09 01:01:10.683: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:01:10.683: INFO @log_variables: valid loss mean: 0.541588 2019-10-09 01:01:10.683: INFO @log_variables: valid age_loss mean: 7.713953 2019-10-09 01:01:10.683: INFO @log_variables: valid gender_loss mean: 0.282179 2019-10-09 01:01:10.683: INFO @log_variables: valid matching_loss nanmean: 0.625343 2019-10-09 01:01:10.683: INFO @log_variables: valid is_face_loss mean: 0.000005 2019-10-09 01:01:10.683: INFO @log_variables: valid age_mae mean: 8.200519 2019-10-09 01:01:10.683: INFO @log_variables: valid gender_accuracy mean: 0.870062 2019-10-09 01:01:10.684: INFO @log_variables: valid positive_distance nanmean: 0.779593 2019-10-09 01:01:10.684: INFO @log_variables: valid negative_distance nanmean: 1.318632 2019-10-09 01:01:10.684: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:01:10.684: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:01:13.006: INFO @metrics_hook: valid matching accuracy: 0.8580679978413384, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:01:13.377: INFO @decay_lr : LR updated to `0.0009655207` 2019-10-09 01:01:13.379: INFO @log_profile : T train: 179.491744 2019-10-09 01:01:13.379: INFO @log_profile : T valid: 8.297618 2019-10-09 01:01:13.379: INFO @log_profile : T read data: 1.482612 2019-10-09 01:01:13.379: INFO @log_profile : T hooks: 3.281954 2019-10-09 01:01:13.379: INFO @main_loop : Epoch 7 done 2019-10-09 01:01:13.379: INFO @main_loop : Training epoch 8 2019-10-09 01:04:22.920: INFO @log_variables: train loss mean: 0.515155 2019-10-09 01:04:22.920: INFO @log_variables: train age_loss mean: 8.038517 2019-10-09 01:04:22.920: INFO @log_variables: train gender_loss mean: 0.243431 2019-10-09 01:04:22.921: INFO @log_variables: train matching_loss nanmean: 0.549684 2019-10-09 01:04:22.921: INFO @log_variables: train is_face_loss mean: 0.000016 2019-10-09 01:04:22.921: INFO @log_variables: train age_mae mean: 8.524899 2019-10-09 01:04:22.921: INFO @log_variables: train gender_accuracy mean: 0.897841 2019-10-09 01:04:22.921: INFO @log_variables: train positive_distance nanmean: 0.831837 2019-10-09 01:04:22.921: INFO @log_variables: train negative_distance nanmean: 1.411409 2019-10-09 01:04:22.921: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:04:22.921: INFO @log_variables: valid loss mean: 0.516980 2019-10-09 01:04:22.921: INFO @log_variables: valid age_loss mean: 6.985425 2019-10-09 01:04:22.921: INFO @log_variables: valid gender_loss mean: 0.301780 2019-10-09 01:04:22.921: INFO @log_variables: valid matching_loss nanmean: 0.602314 2019-10-09 01:04:22.921: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 01:04:22.921: INFO @log_variables: valid age_mae mean: 7.469858 2019-10-09 01:04:22.921: INFO @log_variables: valid gender_accuracy mean: 0.875680 2019-10-09 01:04:22.921: INFO @log_variables: valid positive_distance nanmean: 0.784654 2019-10-09 01:04:22.922: INFO @log_variables: valid negative_distance nanmean: 1.326442 2019-10-09 01:04:22.922: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:04:22.922: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:04:25.567: INFO @metrics_hook: valid matching accuracy: 0.8629849493314146, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:04:25.930: INFO @decay_lr : LR updated to `0.00096069305` 2019-10-09 01:04:25.931: INFO @log_profile : T train: 179.531812 2019-10-09 01:04:25.931: INFO @log_profile : T valid: 8.354846 2019-10-09 01:04:25.931: INFO @log_profile : T read data: 0.991970 2019-10-09 01:04:25.931: INFO @log_profile : T hooks: 3.588586 2019-10-09 01:04:25.931: INFO @main_loop : Epoch 8 done 2019-10-09 01:04:25.931: INFO @main_loop : Training epoch 9 2019-10-09 01:07:36.374: INFO @log_variables: train loss mean: 0.502937 2019-10-09 01:07:36.374: INFO @log_variables: train age_loss mean: 7.875243 2019-10-09 01:07:36.374: INFO @log_variables: train gender_loss mean: 0.232894 2019-10-09 01:07:36.374: INFO @log_variables: train matching_loss nanmean: 0.538673 2019-10-09 01:07:36.374: INFO @log_variables: train is_face_loss mean: 0.000014 2019-10-09 01:07:36.374: INFO @log_variables: train age_mae mean: 8.361094 2019-10-09 01:07:36.375: INFO @log_variables: train gender_accuracy mean: 0.902615 2019-10-09 01:07:36.375: INFO @log_variables: train positive_distance nanmean: 0.830860 2019-10-09 01:07:36.375: INFO @log_variables: train negative_distance nanmean: 1.411357 2019-10-09 01:07:36.375: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:07:36.375: INFO @log_variables: valid loss mean: 0.503400 2019-10-09 01:07:36.375: INFO @log_variables: valid age_loss mean: 7.151602 2019-10-09 01:07:36.375: INFO @log_variables: valid gender_loss mean: 0.262716 2019-10-09 01:07:36.375: INFO @log_variables: valid matching_loss nanmean: 0.582662 2019-10-09 01:07:36.375: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 01:07:36.375: INFO @log_variables: valid age_mae mean: 7.637028 2019-10-09 01:07:36.375: INFO @log_variables: valid gender_accuracy mean: 0.893837 2019-10-09 01:07:36.375: INFO @log_variables: valid positive_distance nanmean: 0.787865 2019-10-09 01:07:36.375: INFO @log_variables: valid negative_distance nanmean: 1.342029 2019-10-09 01:07:36.375: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:07:36.375: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:07:38.584: INFO @metrics_hook: valid matching accuracy: 0.8653234994303531, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:07:38.963: INFO @decay_lr : LR updated to `0.0009558896` 2019-10-09 01:07:39.614: INFO @model : Quantizing and saving the model 2019-10-09 01:07:40.450: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:07:40.457: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:07:40.462: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:07:40.468: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:07:40.473: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:07:40.479: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:07:40.484: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:07:40.490: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:07:40.495: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:07:40.501: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:07:40.506: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:07:40.512: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:07:40.518: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:07:40.524: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:07:40.529: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:07:40.748: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:07:40.755: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:07:40.761: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:07:40.767: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:07:40.772: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:07:40.778: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:07:40.784: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:07:40.790: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:07:40.795: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:07:40.801: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:07:40.807: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:07:40.812: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 01:07:40.820: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 01:07:53.875: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 01:07:54.150: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 01:07:54.170: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 01:07:56.161: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 01:07:56.206: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 01:07:56.211: INFO @log_profile : T train: 179.846314 2019-10-09 01:07:56.211: INFO @log_profile : T valid: 8.521233 2019-10-09 01:07:56.212: INFO @log_profile : T read data: 1.419610 2019-10-09 01:07:56.212: INFO @log_profile : T hooks: 20.407432 2019-10-09 01:07:56.213: INFO @main_loop : Epoch 9 done 2019-10-09 01:07:56.213: INFO @main_loop : Training epoch 10 2019-10-09 01:11:05.840: INFO @log_variables: train loss mean: 0.487350 2019-10-09 01:11:05.840: INFO @log_variables: train age_loss mean: 7.685265 2019-10-09 01:11:05.840: INFO @log_variables: train gender_loss mean: 0.217174 2019-10-09 01:11:05.840: INFO @log_variables: train matching_loss nanmean: 0.525074 2019-10-09 01:11:05.840: INFO @log_variables: train is_face_loss mean: 0.000010 2019-10-09 01:11:05.840: INFO @log_variables: train age_mae mean: 8.171138 2019-10-09 01:11:05.840: INFO @log_variables: train gender_accuracy mean: 0.909991 2019-10-09 01:11:05.840: INFO @log_variables: train positive_distance nanmean: 0.826386 2019-10-09 01:11:05.840: INFO @log_variables: train negative_distance nanmean: 1.411726 2019-10-09 01:11:05.840: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:11:05.841: INFO @log_variables: valid loss mean: 0.495235 2019-10-09 01:11:05.841: INFO @log_variables: valid age_loss mean: 7.012228 2019-10-09 01:11:05.841: INFO @log_variables: valid gender_loss mean: 0.240861 2019-10-09 01:11:05.841: INFO @log_variables: valid matching_loss nanmean: 0.593143 2019-10-09 01:11:05.841: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 01:11:05.841: INFO @log_variables: valid age_mae mean: 7.495918 2019-10-09 01:11:05.841: INFO @log_variables: valid gender_accuracy mean: 0.895257 2019-10-09 01:11:05.841: INFO @log_variables: valid positive_distance nanmean: 0.779549 2019-10-09 01:11:05.841: INFO @log_variables: valid negative_distance nanmean: 1.330551 2019-10-09 01:11:05.841: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:11:05.841: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:11:08.214: INFO @metrics_hook: valid matching accuracy: 0.8675421238831924, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:11:08.589: INFO @decay_lr : LR updated to `0.00095111015` 2019-10-09 01:11:09.246: INFO @model : Quantizing and saving the model 2019-10-09 01:11:10.074: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:11:10.080: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:11:10.085: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:11:10.090: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:11:10.096: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:11:10.101: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:11:10.106: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:11:10.111: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:11:10.116: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:11:10.121: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:11:10.126: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:11:10.131: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:11:10.137: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:11:10.142: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:11:10.147: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:11:10.152: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:11:10.158: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:11:10.163: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:11:10.171: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:11:10.177: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:11:10.182: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:11:10.186: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:11:10.192: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:11:10.197: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:11:10.203: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:11:10.208: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:11:10.214: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 01:11:10.221: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 01:11:21.383: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 01:11:21.633: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 01:11:21.653: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 01:11:23.619: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 01:11:23.656: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 01:11:23.659: INFO @log_profile : T train: 179.185119 2019-10-09 01:11:23.659: INFO @log_profile : T valid: 8.333949 2019-10-09 01:11:23.660: INFO @log_profile : T read data: 1.460227 2019-10-09 01:11:23.660: INFO @log_profile : T hooks: 18.382246 2019-10-09 01:11:23.660: INFO @main_loop : Epoch 10 done 2019-10-09 01:11:23.660: INFO @main_loop : Training epoch 11 2019-10-09 01:14:33.207: INFO @log_variables: train loss mean: 0.482991 2019-10-09 01:14:33.207: INFO @log_variables: train age_loss mean: 7.641072 2019-10-09 01:14:33.207: INFO @log_variables: train gender_loss mean: 0.213586 2019-10-09 01:14:33.207: INFO @log_variables: train matching_loss nanmean: 0.519571 2019-10-09 01:14:33.207: INFO @log_variables: train is_face_loss mean: 0.000009 2019-10-09 01:14:33.207: INFO @log_variables: train age_mae mean: 8.126892 2019-10-09 01:14:33.207: INFO @log_variables: train gender_accuracy mean: 0.911623 2019-10-09 01:14:33.207: INFO @log_variables: train positive_distance nanmean: 0.824508 2019-10-09 01:14:33.207: INFO @log_variables: train negative_distance nanmean: 1.411490 2019-10-09 01:14:33.207: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:14:33.207: INFO @log_variables: valid loss mean: 0.498665 2019-10-09 01:14:33.208: INFO @log_variables: valid age_loss mean: 7.069865 2019-10-09 01:14:33.208: INFO @log_variables: valid gender_loss mean: 0.255647 2019-10-09 01:14:33.208: INFO @log_variables: valid matching_loss nanmean: 0.583226 2019-10-09 01:14:33.208: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 01:14:33.208: INFO @log_variables: valid age_mae mean: 7.552800 2019-10-09 01:14:33.208: INFO @log_variables: valid gender_accuracy mean: 0.890880 2019-10-09 01:14:33.208: INFO @log_variables: valid positive_distance nanmean: 0.788899 2019-10-09 01:14:33.208: INFO @log_variables: valid negative_distance nanmean: 1.338926 2019-10-09 01:14:33.208: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:14:33.208: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:14:36.065: INFO @metrics_hook: valid matching accuracy: 0.8682017149367393, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:14:36.482: INFO @decay_lr : LR updated to `0.0009463546` 2019-10-09 01:14:36.483: INFO @log_profile : T train: 179.372907 2019-10-09 01:14:36.483: INFO @log_profile : T valid: 8.517995 2019-10-09 01:14:36.483: INFO @log_profile : T read data: 0.987678 2019-10-09 01:14:36.483: INFO @log_profile : T hooks: 3.859865 2019-10-09 01:14:36.483: INFO @main_loop : Epoch 11 done 2019-10-09 01:14:36.483: INFO @main_loop : Training epoch 12 2019-10-09 01:17:46.373: INFO @log_variables: train loss mean: 0.465766 2019-10-09 01:17:46.373: INFO @log_variables: train age_loss mean: 7.459913 2019-10-09 01:17:46.373: INFO @log_variables: train gender_loss mean: 0.194271 2019-10-09 01:17:46.373: INFO @log_variables: train matching_loss nanmean: 0.503603 2019-10-09 01:17:46.373: INFO @log_variables: train is_face_loss mean: 0.000008 2019-10-09 01:17:46.373: INFO @log_variables: train age_mae mean: 7.945580 2019-10-09 01:17:46.373: INFO @log_variables: train gender_accuracy mean: 0.922040 2019-10-09 01:17:46.373: INFO @log_variables: train positive_distance nanmean: 0.819156 2019-10-09 01:17:46.373: INFO @log_variables: train negative_distance nanmean: 1.411970 2019-10-09 01:17:46.373: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:17:46.373: INFO @log_variables: valid loss mean: 0.486648 2019-10-09 01:17:46.373: INFO @log_variables: valid age_loss mean: 7.119311 2019-10-09 01:17:46.373: INFO @log_variables: valid gender_loss mean: 0.220448 2019-10-09 01:17:46.373: INFO @log_variables: valid matching_loss nanmean: 0.576228 2019-10-09 01:17:46.373: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 01:17:46.374: INFO @log_variables: valid age_mae mean: 7.604260 2019-10-09 01:17:46.374: INFO @log_variables: valid gender_accuracy mean: 0.906198 2019-10-09 01:17:46.374: INFO @log_variables: valid positive_distance nanmean: 0.785300 2019-10-09 01:17:46.374: INFO @log_variables: valid negative_distance nanmean: 1.342777 2019-10-09 01:17:46.374: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:17:46.374: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:17:48.864: INFO @metrics_hook: valid matching accuracy: 0.8723391497271692, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:17:49.235: INFO @decay_lr : LR updated to `0.0009416228` 2019-10-09 01:17:49.871: INFO @model : Quantizing and saving the model 2019-10-09 01:17:50.971: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:17:50.977: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:17:50.982: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:17:50.987: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:17:50.992: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:17:50.997: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:17:51.004: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:17:51.009: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:17:51.014: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:17:51.020: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:17:51.025: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:17:51.030: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:17:51.036: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:17:51.041: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:17:51.047: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:17:51.052: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:17:51.057: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:17:51.062: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:17:51.068: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:17:51.073: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:17:51.078: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:17:51.083: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:17:51.089: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:17:51.094: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:17:51.100: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:17:51.106: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:17:51.111: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 01:17:51.119: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 01:18:04.240: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 01:18:04.517: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 01:18:04.537: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 01:18:06.525: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 01:18:06.574: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 01:18:06.579: INFO @log_profile : T train: 179.341515 2019-10-09 01:18:06.580: INFO @log_profile : T valid: 8.525690 2019-10-09 01:18:06.580: INFO @log_profile : T read data: 1.354643 2019-10-09 01:18:06.581: INFO @log_profile : T hooks: 20.787364 2019-10-09 01:18:06.581: INFO @main_loop : Epoch 12 done 2019-10-09 01:18:06.581: INFO @main_loop : Training epoch 13 2019-10-09 01:21:16.708: INFO @log_variables: train loss mean: 0.460085 2019-10-09 01:21:16.709: INFO @log_variables: train age_loss mean: 7.337389 2019-10-09 01:21:16.709: INFO @log_variables: train gender_loss mean: 0.192661 2019-10-09 01:21:16.709: INFO @log_variables: train matching_loss nanmean: 0.499857 2019-10-09 01:21:16.709: INFO @log_variables: train is_face_loss mean: 0.000006 2019-10-09 01:21:16.709: INFO @log_variables: train age_mae mean: 7.822579 2019-10-09 01:21:16.709: INFO @log_variables: train gender_accuracy mean: 0.921582 2019-10-09 01:21:16.709: INFO @log_variables: train positive_distance nanmean: 0.819424 2019-10-09 01:21:16.709: INFO @log_variables: train negative_distance nanmean: 1.411682 2019-10-09 01:21:16.709: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:21:16.709: INFO @log_variables: valid loss mean: 0.493434 2019-10-09 01:21:16.709: INFO @log_variables: valid age_loss mean: 7.135378 2019-10-09 01:21:16.709: INFO @log_variables: valid gender_loss mean: 0.244463 2019-10-09 01:21:16.709: INFO @log_variables: valid matching_loss nanmean: 0.571646 2019-10-09 01:21:16.709: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 01:21:16.709: INFO @log_variables: valid age_mae mean: 7.618933 2019-10-09 01:21:16.709: INFO @log_variables: valid gender_accuracy mean: 0.895020 2019-10-09 01:21:16.709: INFO @log_variables: valid positive_distance nanmean: 0.779312 2019-10-09 01:21:16.709: INFO @log_variables: valid negative_distance nanmean: 1.343062 2019-10-09 01:21:16.710: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:21:16.710: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:21:18.942: INFO @metrics_hook: valid matching accuracy: 0.874018108772561, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:21:19.356: INFO @decay_lr : LR updated to `0.0009369147` 2019-10-09 01:21:19.357: INFO @log_profile : T train: 179.473477 2019-10-09 01:21:19.357: INFO @log_profile : T valid: 8.491162 2019-10-09 01:21:19.357: INFO @log_profile : T read data: 1.471054 2019-10-09 01:21:19.358: INFO @log_profile : T hooks: 3.254158 2019-10-09 01:21:19.358: INFO @main_loop : Epoch 13 done 2019-10-09 01:21:19.358: INFO @main_loop : Training epoch 14 2019-10-09 01:24:29.338: INFO @log_variables: train loss mean: 0.448327 2019-10-09 01:24:29.338: INFO @log_variables: train age_loss mean: 7.197576 2019-10-09 01:24:29.338: INFO @log_variables: train gender_loss mean: 0.182846 2019-10-09 01:24:29.338: INFO @log_variables: train matching_loss nanmean: 0.487205 2019-10-09 01:24:29.338: INFO @log_variables: train is_face_loss mean: 0.000005 2019-10-09 01:24:29.338: INFO @log_variables: train age_mae mean: 7.682308 2019-10-09 01:24:29.338: INFO @log_variables: train gender_accuracy mean: 0.926844 2019-10-09 01:24:29.338: INFO @log_variables: train positive_distance nanmean: 0.813345 2019-10-09 01:24:29.338: INFO @log_variables: train negative_distance nanmean: 1.411655 2019-10-09 01:24:29.338: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:24:29.338: INFO @log_variables: valid loss mean: 0.469287 2019-10-09 01:24:29.338: INFO @log_variables: valid age_loss mean: 6.667818 2019-10-09 01:24:29.338: INFO @log_variables: valid gender_loss mean: 0.228477 2019-10-09 01:24:29.338: INFO @log_variables: valid matching_loss nanmean: 0.559530 2019-10-09 01:24:29.338: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 01:24:29.339: INFO @log_variables: valid age_mae mean: 7.152012 2019-10-09 01:24:29.339: INFO @log_variables: valid gender_accuracy mean: 0.907499 2019-10-09 01:24:29.339: INFO @log_variables: valid positive_distance nanmean: 0.797446 2019-10-09 01:24:29.339: INFO @log_variables: valid negative_distance nanmean: 1.354103 2019-10-09 01:24:29.339: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:24:29.339: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:24:31.333: INFO @metrics_hook: valid matching accuracy: 0.873598369011213, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:24:31.731: INFO @decay_lr : LR updated to `0.00093223015` 2019-10-09 01:24:32.388: INFO @model : Quantizing and saving the model 2019-10-09 01:24:33.199: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:24:33.204: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:24:33.209: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:24:33.214: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:24:33.219: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:24:33.224: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:24:33.229: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:24:33.234: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:24:33.239: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:24:33.244: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:24:33.249: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:24:33.254: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:24:33.259: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:24:33.264: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:24:33.269: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:24:33.274: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:24:33.279: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:24:33.284: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:24:33.289: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:24:33.294: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:24:33.299: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:24:33.305: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:24:33.310: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:24:33.315: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:24:33.320: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:24:33.326: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:24:33.331: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 01:24:33.339: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 01:24:45.623: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 01:24:45.896: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 01:24:45.915: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 01:24:47.922: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 01:24:47.963: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 01:24:47.968: INFO @log_profile : T train: 179.775105 2019-10-09 01:24:47.968: INFO @log_profile : T valid: 8.326585 2019-10-09 01:24:47.968: INFO @log_profile : T read data: 1.211299 2019-10-09 01:24:47.968: INFO @log_profile : T hooks: 19.210434 2019-10-09 01:24:47.968: INFO @main_loop : Epoch 14 done 2019-10-09 01:24:47.969: INFO @main_loop : Training epoch 15 2019-10-09 01:27:57.553: INFO @log_variables: train loss mean: 0.439073 2019-10-09 01:27:57.554: INFO @log_variables: train age_loss mean: 7.107176 2019-10-09 01:27:57.554: INFO @log_variables: train gender_loss mean: 0.172256 2019-10-09 01:27:57.554: INFO @log_variables: train matching_loss nanmean: 0.478148 2019-10-09 01:27:57.554: INFO @log_variables: train is_face_loss mean: 0.000004 2019-10-09 01:27:57.554: INFO @log_variables: train age_mae mean: 7.591735 2019-10-09 01:27:57.554: INFO @log_variables: train gender_accuracy mean: 0.931612 2019-10-09 01:27:57.554: INFO @log_variables: train positive_distance nanmean: 0.811375 2019-10-09 01:27:57.554: INFO @log_variables: train negative_distance nanmean: 1.411753 2019-10-09 01:27:57.554: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:27:57.554: INFO @log_variables: valid loss mean: 0.488101 2019-10-09 01:27:57.554: INFO @log_variables: valid age_loss mean: 7.010596 2019-10-09 01:27:57.554: INFO @log_variables: valid gender_loss mean: 0.255796 2019-10-09 01:27:57.555: INFO @log_variables: valid matching_loss nanmean: 0.556256 2019-10-09 01:27:57.555: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 01:27:57.555: INFO @log_variables: valid age_mae mean: 7.495263 2019-10-09 01:27:57.555: INFO @log_variables: valid gender_accuracy mean: 0.898864 2019-10-09 01:27:57.555: INFO @log_variables: valid positive_distance nanmean: 0.781085 2019-10-09 01:27:57.555: INFO @log_variables: valid negative_distance nanmean: 1.357865 2019-10-09 01:27:57.555: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:27:57.555: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:28:00.551: INFO @metrics_hook: valid matching accuracy: 0.873418480542064, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:28:00.947: INFO @decay_lr : LR updated to `0.000927569` 2019-10-09 01:28:00.949: INFO @log_profile : T train: 179.130792 2019-10-09 01:28:00.949: INFO @log_profile : T valid: 8.278093 2019-10-09 01:28:00.949: INFO @log_profile : T read data: 1.463237 2019-10-09 01:28:00.949: INFO @log_profile : T hooks: 4.020523 2019-10-09 01:28:00.949: INFO @main_loop : Epoch 15 done 2019-10-09 01:28:00.949: INFO @main_loop : Training epoch 16 2019-10-09 01:31:09.196: INFO @log_variables: train loss mean: 0.431948 2019-10-09 01:31:09.197: INFO @log_variables: train age_loss mean: 6.996613 2019-10-09 01:31:09.197: INFO @log_variables: train gender_loss mean: 0.164958 2019-10-09 01:31:09.197: INFO @log_variables: train matching_loss nanmean: 0.474415 2019-10-09 01:31:09.197: INFO @log_variables: train is_face_loss mean: 0.000004 2019-10-09 01:31:09.197: INFO @log_variables: train age_mae mean: 7.480881 2019-10-09 01:31:09.197: INFO @log_variables: train gender_accuracy mean: 0.934259 2019-10-09 01:31:09.197: INFO @log_variables: train positive_distance nanmean: 0.810539 2019-10-09 01:31:09.197: INFO @log_variables: train negative_distance nanmean: 1.411899 2019-10-09 01:31:09.197: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:31:09.197: INFO @log_variables: valid loss mean: 0.475232 2019-10-09 01:31:09.197: INFO @log_variables: valid age_loss mean: 7.004312 2019-10-09 01:31:09.197: INFO @log_variables: valid gender_loss mean: 0.214220 2019-10-09 01:31:09.197: INFO @log_variables: valid matching_loss nanmean: 0.558569 2019-10-09 01:31:09.197: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 01:31:09.197: INFO @log_variables: valid age_mae mean: 7.488489 2019-10-09 01:31:09.197: INFO @log_variables: valid gender_accuracy mean: 0.912113 2019-10-09 01:31:09.197: INFO @log_variables: valid positive_distance nanmean: 0.783076 2019-10-09 01:31:09.197: INFO @log_variables: valid negative_distance nanmean: 1.351839 2019-10-09 01:31:09.197: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:31:09.198: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:31:11.615: INFO @metrics_hook: valid matching accuracy: 0.877316064040295, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:31:12.042: INFO @decay_lr : LR updated to `0.0009229311` 2019-10-09 01:31:12.704: INFO @model : Quantizing and saving the model 2019-10-09 01:31:13.892: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:31:13.898: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:31:13.903: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:31:13.909: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:31:13.914: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:31:13.919: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:31:13.925: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:31:13.930: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:31:13.936: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:31:13.941: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:31:13.947: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:31:13.952: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:31:13.958: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:31:13.964: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:31:13.969: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:31:13.975: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:31:13.980: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:31:13.986: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:31:13.991: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:31:13.996: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:31:14.002: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:31:14.009: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:31:14.015: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:31:14.020: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:31:14.025: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 01:31:14.031: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 01:31:14.036: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 01:31:14.043: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 01:31:27.945: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 01:31:28.211: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 01:31:28.230: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 01:31:30.103: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 01:31:30.146: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 01:31:30.150: INFO @log_profile : T train: 178.358480 2019-10-09 01:31:30.151: INFO @log_profile : T valid: 8.271632 2019-10-09 01:31:30.152: INFO @log_profile : T read data: 0.974322 2019-10-09 01:31:30.152: INFO @log_profile : T hooks: 21.512852 2019-10-09 01:31:30.152: INFO @main_loop : Epoch 16 done 2019-10-09 01:31:30.152: INFO @main_loop : Training epoch 17 2019-10-09 01:34:39.087: INFO @log_variables: train loss mean: 0.428245 2019-10-09 01:34:39.087: INFO @log_variables: train age_loss mean: 6.933735 2019-10-09 01:34:39.087: INFO @log_variables: train gender_loss mean: 0.163135 2019-10-09 01:34:39.087: INFO @log_variables: train matching_loss nanmean: 0.471049 2019-10-09 01:34:39.087: INFO @log_variables: train is_face_loss mean: 0.000003 2019-10-09 01:34:39.087: INFO @log_variables: train age_mae mean: 7.418149 2019-10-09 01:34:39.087: INFO @log_variables: train gender_accuracy mean: 0.934997 2019-10-09 01:34:39.087: INFO @log_variables: train positive_distance nanmean: 0.809941 2019-10-09 01:34:39.087: INFO @log_variables: train negative_distance nanmean: 1.411886 2019-10-09 01:34:39.087: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:34:39.087: INFO @log_variables: valid loss mean: 0.478124 2019-10-09 01:34:39.087: INFO @log_variables: valid age_loss mean: 6.871722 2019-10-09 01:34:39.087: INFO @log_variables: valid gender_loss mean: 0.233916 2019-10-09 01:34:39.087: INFO @log_variables: valid matching_loss nanmean: 0.561097 2019-10-09 01:34:39.087: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 01:34:39.087: INFO @log_variables: valid age_mae mean: 7.355894 2019-10-09 01:34:39.088: INFO @log_variables: valid gender_accuracy mean: 0.903478 2019-10-09 01:34:39.088: INFO @log_variables: valid positive_distance nanmean: 0.780769 2019-10-09 01:34:39.088: INFO @log_variables: valid negative_distance nanmean: 1.345002 2019-10-09 01:34:39.088: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:34:39.088: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:34:41.531: INFO @metrics_hook: valid matching accuracy: 0.875037476764406, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:34:41.942: INFO @decay_lr : LR updated to `0.0009183165` 2019-10-09 01:34:41.943: INFO @log_profile : T train: 178.495474 2019-10-09 01:34:41.943: INFO @log_profile : T valid: 8.298123 2019-10-09 01:34:41.943: INFO @log_profile : T read data: 1.464281 2019-10-09 01:34:41.943: INFO @log_profile : T hooks: 3.446877 2019-10-09 01:34:41.943: INFO @main_loop : Epoch 17 done 2019-10-09 01:34:41.943: INFO @main_loop : Training epoch 18 2019-10-09 01:37:51.800: INFO @log_variables: train loss mean: 0.419545 2019-10-09 01:37:51.800: INFO @log_variables: train age_loss mean: 6.793315 2019-10-09 01:37:51.800: INFO @log_variables: train gender_loss mean: 0.157304 2019-10-09 01:37:51.800: INFO @log_variables: train matching_loss nanmean: 0.463951 2019-10-09 01:37:51.801: INFO @log_variables: train is_face_loss mean: 0.000003 2019-10-09 01:37:51.801: INFO @log_variables: train age_mae mean: 7.277258 2019-10-09 01:37:51.801: INFO @log_variables: train gender_accuracy mean: 0.937626 2019-10-09 01:37:51.801: INFO @log_variables: train positive_distance nanmean: 0.806571 2019-10-09 01:37:51.801: INFO @log_variables: train negative_distance nanmean: 1.411646 2019-10-09 01:37:51.801: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:37:51.801: INFO @log_variables: valid loss mean: 0.467269 2019-10-09 01:37:51.801: INFO @log_variables: valid age_loss mean: 6.640812 2019-10-09 01:37:51.801: INFO @log_variables: valid gender_loss mean: 0.233565 2019-10-09 01:37:51.801: INFO @log_variables: valid matching_loss nanmean: 0.550886 2019-10-09 01:37:51.801: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 01:37:51.801: INFO @log_variables: valid age_mae mean: 7.124000 2019-10-09 01:37:51.801: INFO @log_variables: valid gender_accuracy mean: 0.907736 2019-10-09 01:37:51.801: INFO @log_variables: valid positive_distance nanmean: 0.789129 2019-10-09 01:37:51.801: INFO @log_variables: valid negative_distance nanmean: 1.354565 2019-10-09 01:37:51.801: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:37:51.802: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:37:54.334: INFO @metrics_hook: valid matching accuracy: 0.8762367332254003, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:37:54.740: INFO @decay_lr : LR updated to `0.0009137249` 2019-10-09 01:37:54.741: INFO @log_profile : T train: 179.407356 2019-10-09 01:37:54.741: INFO @log_profile : T valid: 8.367105 2019-10-09 01:37:54.741: INFO @log_profile : T read data: 1.419069 2019-10-09 01:37:54.741: INFO @log_profile : T hooks: 3.517915 2019-10-09 01:37:54.741: INFO @main_loop : Epoch 18 done 2019-10-09 01:37:54.741: INFO @main_loop : Training epoch 19 2019-10-09 01:41:03.603: INFO @log_variables: train loss mean: 0.411845 2019-10-09 01:41:03.603: INFO @log_variables: train age_loss mean: 6.725431 2019-10-09 01:41:03.603: INFO @log_variables: train gender_loss mean: 0.146535 2019-10-09 01:41:03.604: INFO @log_variables: train matching_loss nanmean: 0.457639 2019-10-09 01:41:03.604: INFO @log_variables: train is_face_loss mean: 0.000002 2019-10-09 01:41:03.604: INFO @log_variables: train age_mae mean: 7.209267 2019-10-09 01:41:03.604: INFO @log_variables: train gender_accuracy mean: 0.942123 2019-10-09 01:41:03.604: INFO @log_variables: train positive_distance nanmean: 0.804654 2019-10-09 01:41:03.604: INFO @log_variables: train negative_distance nanmean: 1.411865 2019-10-09 01:41:03.604: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:41:03.604: INFO @log_variables: valid loss mean: 0.474906 2019-10-09 01:41:03.604: INFO @log_variables: valid age_loss mean: 6.756788 2019-10-09 01:41:03.604: INFO @log_variables: valid gender_loss mean: 0.246134 2019-10-09 01:41:03.604: INFO @log_variables: valid matching_loss nanmean: 0.550394 2019-10-09 01:41:03.604: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 01:41:03.604: INFO @log_variables: valid age_mae mean: 7.240510 2019-10-09 01:41:03.604: INFO @log_variables: valid gender_accuracy mean: 0.903064 2019-10-09 01:41:03.604: INFO @log_variables: valid positive_distance nanmean: 0.786384 2019-10-09 01:41:03.604: INFO @log_variables: valid negative_distance nanmean: 1.354151 2019-10-09 01:41:03.604: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:41:03.604: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:41:05.973: INFO @metrics_hook: valid matching accuracy: 0.8765365473406488, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:41:06.411: INFO @decay_lr : LR updated to `0.0009091563` 2019-10-09 01:41:06.412: INFO @log_profile : T train: 178.873213 2019-10-09 01:41:06.412: INFO @log_profile : T valid: 8.324365 2019-10-09 01:41:06.412: INFO @log_profile : T read data: 0.993092 2019-10-09 01:41:06.412: INFO @log_profile : T hooks: 3.395875 2019-10-09 01:41:06.412: INFO @main_loop : Epoch 19 done 2019-10-09 01:41:06.412: INFO @main_loop : Training epoch 20 2019-10-09 01:44:15.113: INFO @log_variables: train loss mean: 0.410085 2019-10-09 01:44:15.113: INFO @log_variables: train age_loss mean: 6.709694 2019-10-09 01:44:15.113: INFO @log_variables: train gender_loss mean: 0.145384 2019-10-09 01:44:15.113: INFO @log_variables: train matching_loss nanmean: 0.454910 2019-10-09 01:44:15.114: INFO @log_variables: train is_face_loss mean: 0.000002 2019-10-09 01:44:15.114: INFO @log_variables: train age_mae mean: 7.193532 2019-10-09 01:44:15.114: INFO @log_variables: train gender_accuracy mean: 0.943432 2019-10-09 01:44:15.114: INFO @log_variables: train positive_distance nanmean: 0.804915 2019-10-09 01:44:15.114: INFO @log_variables: train negative_distance nanmean: 1.411601 2019-10-09 01:44:15.114: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:44:15.114: INFO @log_variables: valid loss mean: 0.461568 2019-10-09 01:44:15.114: INFO @log_variables: valid age_loss mean: 6.602052 2019-10-09 01:44:15.114: INFO @log_variables: valid gender_loss mean: 0.218924 2019-10-09 01:44:15.114: INFO @log_variables: valid matching_loss nanmean: 0.551730 2019-10-09 01:44:15.114: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 01:44:15.114: INFO @log_variables: valid age_mae mean: 7.086964 2019-10-09 01:44:15.114: INFO @log_variables: valid gender_accuracy mean: 0.911639 2019-10-09 01:44:15.114: INFO @log_variables: valid positive_distance nanmean: 0.792402 2019-10-09 01:44:15.114: INFO @log_variables: valid negative_distance nanmean: 1.357924 2019-10-09 01:44:15.114: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:44:15.114: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:44:17.783: INFO @metrics_hook: valid matching accuracy: 0.8789950230856869, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:44:18.204: INFO @decay_lr : LR updated to `0.0009046105` 2019-10-09 01:44:18.206: INFO @log_profile : T train: 178.370421 2019-10-09 01:44:18.206: INFO @log_profile : T valid: 8.319470 2019-10-09 01:44:18.206: INFO @log_profile : T read data: 1.364833 2019-10-09 01:44:18.206: INFO @log_profile : T hooks: 3.654928 2019-10-09 01:44:18.206: INFO @main_loop : Epoch 20 done 2019-10-09 01:44:18.206: INFO @main_loop : Training epoch 21 2019-10-09 01:47:27.491: INFO @log_variables: train loss mean: 0.404274 2019-10-09 01:47:27.492: INFO @log_variables: train age_loss mean: 6.568710 2019-10-09 01:47:27.492: INFO @log_variables: train gender_loss mean: 0.144543 2019-10-09 01:47:27.492: INFO @log_variables: train matching_loss nanmean: 0.451832 2019-10-09 01:47:27.492: INFO @log_variables: train is_face_loss mean: 0.000002 2019-10-09 01:47:27.492: INFO @log_variables: train age_mae mean: 7.052189 2019-10-09 01:47:27.492: INFO @log_variables: train gender_accuracy mean: 0.943826 2019-10-09 01:47:27.492: INFO @log_variables: train positive_distance nanmean: 0.803112 2019-10-09 01:47:27.492: INFO @log_variables: train negative_distance nanmean: 1.411695 2019-10-09 01:47:27.492: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:47:27.492: INFO @log_variables: valid loss mean: 0.467340 2019-10-09 01:47:27.492: INFO @log_variables: valid age_loss mean: 6.836041 2019-10-09 01:47:27.492: INFO @log_variables: valid gender_loss mean: 0.225642 2019-10-09 01:47:27.492: INFO @log_variables: valid matching_loss nanmean: 0.539505 2019-10-09 01:47:27.492: INFO @log_variables: valid is_face_loss mean: 0.000002 2019-10-09 01:47:27.492: INFO @log_variables: valid age_mae mean: 7.319466 2019-10-09 01:47:27.492: INFO @log_variables: valid gender_accuracy mean: 0.910634 2019-10-09 01:47:27.492: INFO @log_variables: valid positive_distance nanmean: 0.791115 2019-10-09 01:47:27.492: INFO @log_variables: valid negative_distance nanmean: 1.356974 2019-10-09 01:47:27.492: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:47:27.493: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:47:29.784: INFO @metrics_hook: valid matching accuracy: 0.8819931642381723, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:47:30.194: INFO @decay_lr : LR updated to `0.00090008747` 2019-10-09 01:47:30.196: INFO @log_profile : T train: 178.826742 2019-10-09 01:47:30.196: INFO @log_profile : T valid: 8.308625 2019-10-09 01:47:30.196: INFO @log_profile : T read data: 1.495692 2019-10-09 01:47:30.196: INFO @log_profile : T hooks: 3.274051 2019-10-09 01:47:30.196: INFO @main_loop : Epoch 21 done 2019-10-09 01:47:30.196: INFO @main_loop : Training epoch 22 2019-10-09 01:50:38.619: INFO @log_variables: train loss mean: 0.399264 2019-10-09 01:50:38.619: INFO @log_variables: train age_loss mean: 6.532030 2019-10-09 01:50:38.619: INFO @log_variables: train gender_loss mean: 0.137738 2019-10-09 01:50:38.619: INFO @log_variables: train matching_loss nanmean: 0.446775 2019-10-09 01:50:38.619: INFO @log_variables: train is_face_loss mean: 0.000002 2019-10-09 01:50:38.619: INFO @log_variables: train age_mae mean: 7.015307 2019-10-09 01:50:38.619: INFO @log_variables: train gender_accuracy mean: 0.946674 2019-10-09 01:50:38.619: INFO @log_variables: train positive_distance nanmean: 0.801624 2019-10-09 01:50:38.619: INFO @log_variables: train negative_distance nanmean: 1.411608 2019-10-09 01:50:38.620: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:50:38.620: INFO @log_variables: valid loss mean: 0.472817 2019-10-09 01:50:38.620: INFO @log_variables: valid age_loss mean: 6.787570 2019-10-09 01:50:38.620: INFO @log_variables: valid gender_loss mean: 0.246837 2019-10-09 01:50:38.620: INFO @log_variables: valid matching_loss nanmean: 0.540136 2019-10-09 01:50:38.620: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 01:50:38.620: INFO @log_variables: valid age_mae mean: 7.271413 2019-10-09 01:50:38.620: INFO @log_variables: valid gender_accuracy mean: 0.902295 2019-10-09 01:50:38.620: INFO @log_variables: valid positive_distance nanmean: 0.787983 2019-10-09 01:50:38.620: INFO @log_variables: valid negative_distance nanmean: 1.355939 2019-10-09 01:50:38.620: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:50:38.620: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:50:40.894: INFO @metrics_hook: valid matching accuracy: 0.8809737962463273, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:50:41.300: INFO @decay_lr : LR updated to `0.000895587` 2019-10-09 01:50:41.302: INFO @log_profile : T train: 178.381178 2019-10-09 01:50:41.302: INFO @log_profile : T valid: 8.340989 2019-10-09 01:50:41.302: INFO @log_profile : T read data: 1.024440 2019-10-09 01:50:41.302: INFO @log_profile : T hooks: 3.272758 2019-10-09 01:50:41.302: INFO @main_loop : Epoch 22 done 2019-10-09 01:50:41.302: INFO @main_loop : Training epoch 23 2019-10-09 01:53:50.447: INFO @log_variables: train loss mean: 0.394816 2019-10-09 01:53:50.447: INFO @log_variables: train age_loss mean: 6.473001 2019-10-09 01:53:50.447: INFO @log_variables: train gender_loss mean: 0.130935 2019-10-09 01:53:50.447: INFO @log_variables: train matching_loss nanmean: 0.445692 2019-10-09 01:53:50.447: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 01:53:50.447: INFO @log_variables: train age_mae mean: 6.956261 2019-10-09 01:53:50.447: INFO @log_variables: train gender_accuracy mean: 0.948816 2019-10-09 01:53:50.447: INFO @log_variables: train positive_distance nanmean: 0.801295 2019-10-09 01:53:50.447: INFO @log_variables: train negative_distance nanmean: 1.411798 2019-10-09 01:53:50.448: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:53:50.448: INFO @log_variables: valid loss mean: 0.479264 2019-10-09 01:53:50.448: INFO @log_variables: valid age_loss mean: 6.947985 2019-10-09 01:53:50.448: INFO @log_variables: valid gender_loss mean: 0.236383 2019-10-09 01:53:50.448: INFO @log_variables: valid matching_loss nanmean: 0.554536 2019-10-09 01:53:50.448: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 01:53:50.448: INFO @log_variables: valid age_mae mean: 7.432804 2019-10-09 01:53:50.448: INFO @log_variables: valid gender_accuracy mean: 0.904483 2019-10-09 01:53:50.448: INFO @log_variables: valid positive_distance nanmean: 0.783355 2019-10-09 01:53:50.448: INFO @log_variables: valid negative_distance nanmean: 1.347708 2019-10-09 01:53:50.448: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:53:50.448: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:53:52.521: INFO @metrics_hook: valid matching accuracy: 0.878755171793488, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:53:52.939: INFO @decay_lr : LR updated to `0.0008911091` 2019-10-09 01:53:52.940: INFO @log_profile : T train: 178.737136 2019-10-09 01:53:52.940: INFO @log_profile : T valid: 8.379884 2019-10-09 01:53:52.940: INFO @log_profile : T read data: 1.392556 2019-10-09 01:53:52.940: INFO @log_profile : T hooks: 3.042743 2019-10-09 01:53:52.940: INFO @main_loop : Epoch 23 done 2019-10-09 01:53:52.940: INFO @main_loop : Training epoch 24 2019-10-09 01:57:01.363: INFO @log_variables: train loss mean: 0.391413 2019-10-09 01:57:01.363: INFO @log_variables: train age_loss mean: 6.392697 2019-10-09 01:57:01.363: INFO @log_variables: train gender_loss mean: 0.131359 2019-10-09 01:57:01.363: INFO @log_variables: train matching_loss nanmean: 0.442751 2019-10-09 01:57:01.363: INFO @log_variables: train is_face_loss mean: 0.000002 2019-10-09 01:57:01.363: INFO @log_variables: train age_mae mean: 6.875343 2019-10-09 01:57:01.363: INFO @log_variables: train gender_accuracy mean: 0.948756 2019-10-09 01:57:01.363: INFO @log_variables: train positive_distance nanmean: 0.800752 2019-10-09 01:57:01.363: INFO @log_variables: train negative_distance nanmean: 1.411776 2019-10-09 01:57:01.363: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 01:57:01.363: INFO @log_variables: valid loss mean: 0.476378 2019-10-09 01:57:01.363: INFO @log_variables: valid age_loss mean: 6.866021 2019-10-09 01:57:01.363: INFO @log_variables: valid gender_loss mean: 0.244802 2019-10-09 01:57:01.363: INFO @log_variables: valid matching_loss nanmean: 0.545365 2019-10-09 01:57:01.363: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 01:57:01.364: INFO @log_variables: valid age_mae mean: 7.351241 2019-10-09 01:57:01.364: INFO @log_variables: valid gender_accuracy mean: 0.905193 2019-10-09 01:57:01.364: INFO @log_variables: valid positive_distance nanmean: 0.780937 2019-10-09 01:57:01.364: INFO @log_variables: valid negative_distance nanmean: 1.351248 2019-10-09 01:57:01.364: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 01:57:01.364: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 01:57:03.784: INFO @metrics_hook: valid matching accuracy: 0.8816933501229238, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 01:57:04.225: INFO @decay_lr : LR updated to `0.00088665355` 2019-10-09 01:57:04.226: INFO @log_profile : T train: 178.475266 2019-10-09 01:57:04.226: INFO @log_profile : T valid: 8.288657 2019-10-09 01:57:04.226: INFO @log_profile : T read data: 1.012025 2019-10-09 01:57:04.226: INFO @log_profile : T hooks: 3.425244 2019-10-09 01:57:04.226: INFO @main_loop : Epoch 24 done 2019-10-09 01:57:04.227: INFO @main_loop : Training epoch 25 2019-10-09 02:00:13.034: INFO @log_variables: train loss mean: 0.383334 2019-10-09 02:00:13.034: INFO @log_variables: train age_loss mean: 6.312034 2019-10-09 02:00:13.034: INFO @log_variables: train gender_loss mean: 0.120981 2019-10-09 02:00:13.034: INFO @log_variables: train matching_loss nanmean: 0.436149 2019-10-09 02:00:13.034: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 02:00:13.034: INFO @log_variables: train age_mae mean: 6.794684 2019-10-09 02:00:13.034: INFO @log_variables: train gender_accuracy mean: 0.953575 2019-10-09 02:00:13.034: INFO @log_variables: train positive_distance nanmean: 0.797168 2019-10-09 02:00:13.034: INFO @log_variables: train negative_distance nanmean: 1.411584 2019-10-09 02:00:13.034: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:00:13.034: INFO @log_variables: valid loss mean: 0.473713 2019-10-09 02:00:13.034: INFO @log_variables: valid age_loss mean: 6.800358 2019-10-09 02:00:13.034: INFO @log_variables: valid gender_loss mean: 0.236787 2019-10-09 02:00:13.034: INFO @log_variables: valid matching_loss nanmean: 0.551687 2019-10-09 02:00:13.034: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 02:00:13.035: INFO @log_variables: valid age_mae mean: 7.285089 2019-10-09 02:00:13.035: INFO @log_variables: valid gender_accuracy mean: 0.913236 2019-10-09 02:00:13.035: INFO @log_variables: valid positive_distance nanmean: 0.774589 2019-10-09 02:00:13.035: INFO @log_variables: valid negative_distance nanmean: 1.348445 2019-10-09 02:00:13.035: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:00:13.035: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:00:15.446: INFO @metrics_hook: valid matching accuracy: 0.8801942795466811, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:00:15.850: INFO @decay_lr : LR updated to `0.0008822203` 2019-10-09 02:00:16.529: INFO @model : Quantizing and saving the model 2019-10-09 02:00:17.384: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:00:17.390: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:00:17.396: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:00:17.401: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:00:17.407: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:00:17.412: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:00:17.418: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:00:17.423: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:00:17.429: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:00:17.434: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:00:17.440: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:00:17.446: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:00:17.451: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:00:17.456: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:00:17.461: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:00:17.466: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:00:17.471: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:00:17.476: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:00:17.482: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:00:17.487: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:00:17.492: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:00:17.497: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:00:17.502: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:00:17.507: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:00:17.512: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:00:17.517: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:00:17.522: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 02:00:17.529: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 02:00:29.772: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 02:00:30.044: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 02:00:30.064: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 02:00:31.992: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 02:00:32.036: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 02:00:32.040: INFO @log_profile : T train: 178.428493 2019-10-09 02:00:32.040: INFO @log_profile : T valid: 8.321743 2019-10-09 02:00:32.040: INFO @log_profile : T read data: 1.404589 2019-10-09 02:00:32.041: INFO @log_profile : T hooks: 19.574057 2019-10-09 02:00:32.041: INFO @main_loop : Epoch 25 done 2019-10-09 02:00:32.041: INFO @main_loop : Training epoch 26 2019-10-09 02:03:40.856: INFO @log_variables: train loss mean: 0.380231 2019-10-09 02:03:40.856: INFO @log_variables: train age_loss mean: 6.241656 2019-10-09 02:03:40.856: INFO @log_variables: train gender_loss mean: 0.121857 2019-10-09 02:03:40.856: INFO @log_variables: train matching_loss nanmean: 0.432693 2019-10-09 02:03:40.856: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 02:03:40.856: INFO @log_variables: train age_mae mean: 6.724329 2019-10-09 02:03:40.856: INFO @log_variables: train gender_accuracy mean: 0.952358 2019-10-09 02:03:40.856: INFO @log_variables: train positive_distance nanmean: 0.794919 2019-10-09 02:03:40.856: INFO @log_variables: train negative_distance nanmean: 1.411862 2019-10-09 02:03:40.856: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:03:40.857: INFO @log_variables: valid loss mean: 0.497976 2019-10-09 02:03:40.857: INFO @log_variables: valid age_loss mean: 6.783634 2019-10-09 02:03:40.857: INFO @log_variables: valid gender_loss mean: 0.300662 2019-10-09 02:03:40.857: INFO @log_variables: valid matching_loss nanmean: 0.564701 2019-10-09 02:03:40.857: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 02:03:40.857: INFO @log_variables: valid age_mae mean: 7.267501 2019-10-09 02:03:40.857: INFO @log_variables: valid gender_accuracy mean: 0.880944 2019-10-09 02:03:40.857: INFO @log_variables: valid positive_distance nanmean: 0.769515 2019-10-09 02:03:40.857: INFO @log_variables: valid negative_distance nanmean: 1.336332 2019-10-09 02:03:40.857: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:03:40.857: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:03:43.176: INFO @metrics_hook: valid matching accuracy: 0.8790549859087365, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:03:43.581: INFO @decay_lr : LR updated to `0.0008778092` 2019-10-09 02:03:43.582: INFO @log_profile : T train: 178.438390 2019-10-09 02:03:43.582: INFO @log_profile : T valid: 8.284153 2019-10-09 02:03:43.582: INFO @log_profile : T read data: 1.429173 2019-10-09 02:03:43.582: INFO @log_profile : T hooks: 3.302682 2019-10-09 02:03:43.582: INFO @main_loop : Epoch 26 done 2019-10-09 02:03:43.582: INFO @main_loop : Training epoch 27 2019-10-09 02:06:51.953: INFO @log_variables: train loss mean: 0.378207 2019-10-09 02:06:51.954: INFO @log_variables: train age_loss mean: 6.227781 2019-10-09 02:06:51.954: INFO @log_variables: train gender_loss mean: 0.119260 2019-10-09 02:06:51.954: INFO @log_variables: train matching_loss nanmean: 0.430403 2019-10-09 02:06:51.954: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 02:06:51.954: INFO @log_variables: train age_mae mean: 6.710412 2019-10-09 02:06:51.954: INFO @log_variables: train gender_accuracy mean: 0.954051 2019-10-09 02:06:51.954: INFO @log_variables: train positive_distance nanmean: 0.795720 2019-10-09 02:06:51.954: INFO @log_variables: train negative_distance nanmean: 1.411661 2019-10-09 02:06:51.954: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:06:51.954: INFO @log_variables: valid loss mean: 0.467982 2019-10-09 02:06:51.954: INFO @log_variables: valid age_loss mean: 6.875109 2019-10-09 02:06:51.954: INFO @log_variables: valid gender_loss mean: 0.221614 2019-10-09 02:06:51.954: INFO @log_variables: valid matching_loss nanmean: 0.541619 2019-10-09 02:06:51.954: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:06:51.954: INFO @log_variables: valid age_mae mean: 7.359112 2019-10-09 02:06:51.954: INFO @log_variables: valid gender_accuracy mean: 0.913828 2019-10-09 02:06:51.954: INFO @log_variables: valid positive_distance nanmean: 0.788537 2019-10-09 02:06:51.954: INFO @log_variables: valid negative_distance nanmean: 1.358004 2019-10-09 02:06:51.954: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:06:51.955: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:06:54.212: INFO @metrics_hook: valid matching accuracy: 0.8819931642381723, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:06:54.622: INFO @decay_lr : LR updated to `0.00087342015` 2019-10-09 02:06:55.527: INFO @model : Quantizing and saving the model 2019-10-09 02:06:56.370: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:06:56.376: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:06:56.381: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:06:56.386: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:06:56.392: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:06:56.397: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:06:56.402: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:06:56.407: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:06:56.412: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:06:56.418: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:06:56.423: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:06:56.428: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:06:56.433: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:06:56.438: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:06:56.444: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:06:56.449: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:06:56.455: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:06:56.460: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:06:56.466: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:06:56.471: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:06:56.477: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:06:56.482: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:06:56.487: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:06:56.493: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:06:56.499: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:06:56.504: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:06:56.510: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 02:06:56.518: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 02:07:09.365: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 02:07:09.633: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 02:07:09.652: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 02:07:11.568: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 02:07:11.612: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 02:07:11.616: INFO @log_profile : T train: 178.367713 2019-10-09 02:07:11.616: INFO @log_profile : T valid: 8.360239 2019-10-09 02:07:11.617: INFO @log_profile : T read data: 0.991527 2019-10-09 02:07:11.617: INFO @log_profile : T hooks: 20.227520 2019-10-09 02:07:11.617: INFO @main_loop : Epoch 27 done 2019-10-09 02:07:11.617: INFO @main_loop : Training epoch 28 2019-10-09 02:10:20.780: INFO @log_variables: train loss mean: 0.374182 2019-10-09 02:10:20.780: INFO @log_variables: train age_loss mean: 6.169041 2019-10-09 02:10:20.780: INFO @log_variables: train gender_loss mean: 0.113814 2019-10-09 02:10:20.780: INFO @log_variables: train matching_loss nanmean: 0.429246 2019-10-09 02:10:20.780: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 02:10:20.780: INFO @log_variables: train age_mae mean: 6.651157 2019-10-09 02:10:20.780: INFO @log_variables: train gender_accuracy mean: 0.956899 2019-10-09 02:10:20.780: INFO @log_variables: train positive_distance nanmean: 0.793718 2019-10-09 02:10:20.781: INFO @log_variables: train negative_distance nanmean: 1.411539 2019-10-09 02:10:20.781: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:10:20.781: INFO @log_variables: valid loss mean: 0.455269 2019-10-09 02:10:20.781: INFO @log_variables: valid age_loss mean: 6.507181 2019-10-09 02:10:20.781: INFO @log_variables: valid gender_loss mean: 0.228703 2019-10-09 02:10:20.781: INFO @log_variables: valid matching_loss nanmean: 0.531914 2019-10-09 02:10:20.781: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:10:20.781: INFO @log_variables: valid age_mae mean: 6.990983 2019-10-09 02:10:20.781: INFO @log_variables: valid gender_accuracy mean: 0.912113 2019-10-09 02:10:20.781: INFO @log_variables: valid positive_distance nanmean: 0.791722 2019-10-09 02:10:20.781: INFO @log_variables: valid negative_distance nanmean: 1.365746 2019-10-09 02:10:20.781: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:10:20.781: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:10:23.134: INFO @metrics_hook: valid matching accuracy: 0.8818132757690232, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:10:23.523: INFO @decay_lr : LR updated to `0.00086905307` 2019-10-09 02:10:23.525: INFO @log_profile : T train: 178.811685 2019-10-09 02:10:23.525: INFO @log_profile : T valid: 8.300295 2019-10-09 02:10:23.525: INFO @log_profile : T read data: 1.393278 2019-10-09 02:10:23.525: INFO @log_profile : T hooks: 3.317448 2019-10-09 02:10:23.525: INFO @main_loop : Epoch 28 done 2019-10-09 02:10:23.525: INFO @main_loop : Training epoch 29 2019-10-09 02:13:33.704: INFO @log_variables: train loss mean: 0.367928 2019-10-09 02:13:33.705: INFO @log_variables: train age_loss mean: 6.090776 2019-10-09 02:13:33.705: INFO @log_variables: train gender_loss mean: 0.112080 2019-10-09 02:13:33.705: INFO @log_variables: train matching_loss nanmean: 0.419418 2019-10-09 02:13:33.705: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 02:13:33.705: INFO @log_variables: train age_mae mean: 6.572932 2019-10-09 02:13:33.705: INFO @log_variables: train gender_accuracy mean: 0.957231 2019-10-09 02:13:33.705: INFO @log_variables: train positive_distance nanmean: 0.789507 2019-10-09 02:13:33.705: INFO @log_variables: train negative_distance nanmean: 1.411698 2019-10-09 02:13:33.705: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:13:33.705: INFO @log_variables: valid loss mean: 0.473027 2019-10-09 02:13:33.705: INFO @log_variables: valid age_loss mean: 6.875621 2019-10-09 02:13:33.705: INFO @log_variables: valid gender_loss mean: 0.224689 2019-10-09 02:13:33.705: INFO @log_variables: valid matching_loss nanmean: 0.554132 2019-10-09 02:13:33.705: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:13:33.705: INFO @log_variables: valid age_mae mean: 7.359781 2019-10-09 02:13:33.705: INFO @log_variables: valid gender_accuracy mean: 0.910575 2019-10-09 02:13:33.705: INFO @log_variables: valid positive_distance nanmean: 0.779226 2019-10-09 02:13:33.705: INFO @log_variables: valid negative_distance nanmean: 1.350427 2019-10-09 02:13:33.705: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:13:33.706: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:13:36.178: INFO @metrics_hook: valid matching accuracy: 0.8784553576782395, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:13:36.573: INFO @decay_lr : LR updated to `0.0008647078` 2019-10-09 02:13:36.575: INFO @log_profile : T train: 179.570879 2019-10-09 02:13:36.575: INFO @log_profile : T valid: 8.572664 2019-10-09 02:13:36.575: INFO @log_profile : T read data: 1.421345 2019-10-09 02:13:36.575: INFO @log_profile : T hooks: 3.398557 2019-10-09 02:13:36.575: INFO @main_loop : Epoch 29 done 2019-10-09 02:13:36.575: INFO @main_loop : Training epoch 30 2019-10-09 02:16:44.989: INFO @log_variables: train loss mean: 0.366881 2019-10-09 02:16:44.989: INFO @log_variables: train age_loss mean: 6.036043 2019-10-09 02:16:44.989: INFO @log_variables: train gender_loss mean: 0.109735 2019-10-09 02:16:44.989: INFO @log_variables: train matching_loss nanmean: 0.423990 2019-10-09 02:16:44.989: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 02:16:44.989: INFO @log_variables: train age_mae mean: 6.517760 2019-10-09 02:16:44.989: INFO @log_variables: train gender_accuracy mean: 0.958659 2019-10-09 02:16:44.989: INFO @log_variables: train positive_distance nanmean: 0.792063 2019-10-09 02:16:44.989: INFO @log_variables: train negative_distance nanmean: 1.411386 2019-10-09 02:16:44.989: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:16:44.990: INFO @log_variables: valid loss mean: 0.460558 2019-10-09 02:16:44.990: INFO @log_variables: valid age_loss mean: 6.539457 2019-10-09 02:16:44.990: INFO @log_variables: valid gender_loss mean: 0.245826 2019-10-09 02:16:44.990: INFO @log_variables: valid matching_loss nanmean: 0.527956 2019-10-09 02:16:44.990: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 02:16:44.990: INFO @log_variables: valid age_mae mean: 7.023621 2019-10-09 02:16:44.990: INFO @log_variables: valid gender_accuracy mean: 0.910279 2019-10-09 02:16:44.990: INFO @log_variables: valid positive_distance nanmean: 0.788798 2019-10-09 02:16:44.990: INFO @log_variables: valid negative_distance nanmean: 1.365898 2019-10-09 02:16:44.990: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:16:44.990: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:16:47.390: INFO @metrics_hook: valid matching accuracy: 0.8833723091683157, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:16:47.810: INFO @decay_lr : LR updated to `0.00086038426` 2019-10-09 02:16:47.811: INFO @log_profile : T train: 178.480441 2019-10-09 02:16:47.811: INFO @log_profile : T valid: 8.291794 2019-10-09 02:16:47.812: INFO @log_profile : T read data: 0.999957 2019-10-09 02:16:47.812: INFO @log_profile : T hooks: 3.379059 2019-10-09 02:16:47.812: INFO @main_loop : Epoch 30 done 2019-10-09 02:16:47.812: INFO @main_loop : Training epoch 31 2019-10-09 02:19:56.940: INFO @log_variables: train loss mean: 0.364717 2019-10-09 02:19:56.940: INFO @log_variables: train age_loss mean: 6.022429 2019-10-09 02:19:56.940: INFO @log_variables: train gender_loss mean: 0.106747 2019-10-09 02:19:56.940: INFO @log_variables: train matching_loss nanmean: 0.421634 2019-10-09 02:19:56.940: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 02:19:56.940: INFO @log_variables: train age_mae mean: 6.504690 2019-10-09 02:19:56.940: INFO @log_variables: train gender_accuracy mean: 0.959859 2019-10-09 02:19:56.940: INFO @log_variables: train positive_distance nanmean: 0.792359 2019-10-09 02:19:56.940: INFO @log_variables: train negative_distance nanmean: 1.411185 2019-10-09 02:19:56.940: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:19:56.940: INFO @log_variables: valid loss mean: 0.453181 2019-10-09 02:19:56.940: INFO @log_variables: valid age_loss mean: 6.439631 2019-10-09 02:19:56.941: INFO @log_variables: valid gender_loss mean: 0.228457 2019-10-09 02:19:56.941: INFO @log_variables: valid matching_loss nanmean: 0.532441 2019-10-09 02:19:56.941: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 02:19:56.941: INFO @log_variables: valid age_mae mean: 6.923260 2019-10-09 02:19:56.941: INFO @log_variables: valid gender_accuracy mean: 0.911344 2019-10-09 02:19:56.941: INFO @log_variables: valid positive_distance nanmean: 0.786624 2019-10-09 02:19:56.941: INFO @log_variables: valid negative_distance nanmean: 1.361075 2019-10-09 02:19:56.941: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:19:56.941: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:19:58.860: INFO @metrics_hook: valid matching accuracy: 0.882892606583918, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:19:59.261: INFO @decay_lr : LR updated to `0.00085608236` 2019-10-09 02:19:59.262: INFO @log_profile : T train: 178.463424 2019-10-09 02:19:59.262: INFO @log_profile : T valid: 8.411043 2019-10-09 02:19:59.262: INFO @log_profile : T read data: 1.603357 2019-10-09 02:19:59.262: INFO @log_profile : T hooks: 2.886914 2019-10-09 02:19:59.262: INFO @main_loop : Epoch 31 done 2019-10-09 02:19:59.263: INFO @main_loop : Training epoch 32 2019-10-09 02:23:09.026: INFO @log_variables: train loss mean: 0.361260 2019-10-09 02:23:09.026: INFO @log_variables: train age_loss mean: 5.976030 2019-10-09 02:23:09.026: INFO @log_variables: train gender_loss mean: 0.103773 2019-10-09 02:23:09.026: INFO @log_variables: train matching_loss nanmean: 0.418529 2019-10-09 02:23:09.026: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 02:23:09.026: INFO @log_variables: train age_mae mean: 6.457827 2019-10-09 02:23:09.026: INFO @log_variables: train gender_accuracy mean: 0.960132 2019-10-09 02:23:09.026: INFO @log_variables: train positive_distance nanmean: 0.789125 2019-10-09 02:23:09.027: INFO @log_variables: train negative_distance nanmean: 1.411412 2019-10-09 02:23:09.027: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:23:09.027: INFO @log_variables: valid loss mean: 0.456973 2019-10-09 02:23:09.027: INFO @log_variables: valid age_loss mean: 6.639103 2019-10-09 02:23:09.027: INFO @log_variables: valid gender_loss mean: 0.218327 2019-10-09 02:23:09.027: INFO @log_variables: valid matching_loss nanmean: 0.534378 2019-10-09 02:23:09.027: INFO @log_variables: valid is_face_loss mean: 0.000001 2019-10-09 02:23:09.027: INFO @log_variables: valid age_mae mean: 7.123246 2019-10-09 02:23:09.027: INFO @log_variables: valid gender_accuracy mean: 0.914064 2019-10-09 02:23:09.027: INFO @log_variables: valid positive_distance nanmean: 0.785872 2019-10-09 02:23:09.027: INFO @log_variables: valid negative_distance nanmean: 1.363943 2019-10-09 02:23:09.027: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:23:09.027: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:23:11.667: INFO @metrics_hook: valid matching accuracy: 0.8801343167236313, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:23:12.057: INFO @decay_lr : LR updated to `0.00085180195` 2019-10-09 02:23:12.740: INFO @model : Quantizing and saving the model 2019-10-09 02:23:13.581: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:23:13.586: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:23:13.592: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:23:13.597: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:23:13.602: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:23:13.607: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:23:13.613: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:23:13.618: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:23:13.623: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:23:13.628: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:23:13.633: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:23:13.639: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:23:13.644: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:23:13.649: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:23:13.654: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:23:13.659: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:23:13.665: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:23:13.671: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:23:13.955: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:23:13.964: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:23:13.970: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:23:13.976: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:23:13.981: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:23:13.987: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:23:13.992: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:23:13.997: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:23:14.003: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 02:23:14.010: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 02:23:28.800: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 02:23:29.236: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 02:23:29.265: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 02:23:31.366: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 02:23:31.418: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 02:23:31.423: INFO @log_profile : T train: 179.639910 2019-10-09 02:23:31.425: INFO @log_profile : T valid: 8.499807 2019-10-09 02:23:31.427: INFO @log_profile : T read data: 0.986380 2019-10-09 02:23:31.427: INFO @log_profile : T hooks: 22.947966 2019-10-09 02:23:31.427: INFO @main_loop : Epoch 32 done 2019-10-09 02:23:31.427: INFO @main_loop : Training epoch 33 2019-10-09 02:26:40.646: INFO @log_variables: train loss mean: 0.359696 2019-10-09 02:26:40.647: INFO @log_variables: train age_loss mean: 5.977410 2019-10-09 02:26:40.647: INFO @log_variables: train gender_loss mean: 0.101808 2019-10-09 02:26:40.647: INFO @log_variables: train matching_loss nanmean: 0.415506 2019-10-09 02:26:40.647: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 02:26:40.647: INFO @log_variables: train age_mae mean: 6.459179 2019-10-09 02:26:40.647: INFO @log_variables: train gender_accuracy mean: 0.960970 2019-10-09 02:26:40.647: INFO @log_variables: train positive_distance nanmean: 0.786888 2019-10-09 02:26:40.647: INFO @log_variables: train negative_distance nanmean: 1.410913 2019-10-09 02:26:40.647: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:26:40.647: INFO @log_variables: valid loss mean: 0.468254 2019-10-09 02:26:40.647: INFO @log_variables: valid age_loss mean: 6.958057 2019-10-09 02:26:40.647: INFO @log_variables: valid gender_loss mean: 0.221992 2019-10-09 02:26:40.647: INFO @log_variables: valid matching_loss nanmean: 0.533790 2019-10-09 02:26:40.647: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:26:40.647: INFO @log_variables: valid age_mae mean: 7.442280 2019-10-09 02:26:40.647: INFO @log_variables: valid gender_accuracy mean: 0.917613 2019-10-09 02:26:40.647: INFO @log_variables: valid positive_distance nanmean: 0.778263 2019-10-09 02:26:40.647: INFO @log_variables: valid negative_distance nanmean: 1.357164 2019-10-09 02:26:40.647: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:26:40.648: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:26:43.201: INFO @metrics_hook: valid matching accuracy: 0.8836721232835641, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:26:43.617: INFO @decay_lr : LR updated to `0.00084754295` 2019-10-09 02:26:44.280: INFO @model : Quantizing and saving the model 2019-10-09 02:26:45.113: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:26:45.119: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:26:45.124: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:26:45.130: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:26:45.136: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:26:45.141: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:26:45.146: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:26:45.151: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:26:45.156: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:26:45.162: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:26:45.167: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:26:45.172: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:26:45.178: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:26:45.184: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:26:45.190: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:26:45.195: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:26:45.201: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:26:45.206: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:26:45.212: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:26:45.218: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:26:45.223: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:26:45.229: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:26:45.235: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:26:45.240: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:26:45.246: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:26:45.251: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:26:45.257: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 02:26:45.264: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 02:26:57.653: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 02:26:57.926: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 02:26:57.945: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 02:26:59.903: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 02:26:59.940: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 02:26:59.943: INFO @log_profile : T train: 178.781052 2019-10-09 02:26:59.944: INFO @log_profile : T valid: 8.290858 2019-10-09 02:26:59.945: INFO @log_profile : T read data: 1.458378 2019-10-09 02:26:59.945: INFO @log_profile : T hooks: 19.900220 2019-10-09 02:26:59.945: INFO @main_loop : Epoch 33 done 2019-10-09 02:26:59.945: INFO @main_loop : Training epoch 34 2019-10-09 02:30:08.662: INFO @log_variables: train loss mean: 0.357043 2019-10-09 02:30:08.662: INFO @log_variables: train age_loss mean: 5.913998 2019-10-09 02:30:08.662: INFO @log_variables: train gender_loss mean: 0.100217 2019-10-09 02:30:08.662: INFO @log_variables: train matching_loss nanmean: 0.415216 2019-10-09 02:30:08.662: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 02:30:08.662: INFO @log_variables: train age_mae mean: 6.395792 2019-10-09 02:30:08.663: INFO @log_variables: train gender_accuracy mean: 0.962133 2019-10-09 02:30:08.663: INFO @log_variables: train positive_distance nanmean: 0.788764 2019-10-09 02:30:08.663: INFO @log_variables: train negative_distance nanmean: 1.410612 2019-10-09 02:30:08.663: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:30:08.663: INFO @log_variables: valid loss mean: 0.463761 2019-10-09 02:30:08.663: INFO @log_variables: valid age_loss mean: 6.518673 2019-10-09 02:30:08.663: INFO @log_variables: valid gender_loss mean: 0.264614 2019-10-09 02:30:08.663: INFO @log_variables: valid matching_loss nanmean: 0.521177 2019-10-09 02:30:08.663: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:30:08.663: INFO @log_variables: valid age_mae mean: 7.002834 2019-10-09 02:30:08.663: INFO @log_variables: valid gender_accuracy mean: 0.907381 2019-10-09 02:30:08.663: INFO @log_variables: valid positive_distance nanmean: 0.786198 2019-10-09 02:30:08.663: INFO @log_variables: valid negative_distance nanmean: 1.367022 2019-10-09 02:30:08.663: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:30:08.663: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:30:10.892: INFO @metrics_hook: valid matching accuracy: 0.8817533129459735, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:30:11.314: INFO @decay_lr : LR updated to `0.00084330526` 2019-10-09 02:30:11.316: INFO @log_profile : T train: 178.263953 2019-10-09 02:30:11.316: INFO @log_profile : T valid: 8.319105 2019-10-09 02:30:11.316: INFO @log_profile : T read data: 1.445680 2019-10-09 02:30:11.316: INFO @log_profile : T hooks: 3.254015 2019-10-09 02:30:11.316: INFO @main_loop : Epoch 34 done 2019-10-09 02:30:11.316: INFO @main_loop : Training epoch 35 2019-10-09 02:33:19.520: INFO @log_variables: train loss mean: 0.355319 2019-10-09 02:33:19.520: INFO @log_variables: train age_loss mean: 5.846451 2019-10-09 02:33:19.521: INFO @log_variables: train gender_loss mean: 0.101284 2019-10-09 02:33:19.521: INFO @log_variables: train matching_loss nanmean: 0.415558 2019-10-09 02:33:19.521: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 02:33:19.521: INFO @log_variables: train age_mae mean: 6.328176 2019-10-09 02:33:19.521: INFO @log_variables: train gender_accuracy mean: 0.961012 2019-10-09 02:33:19.521: INFO @log_variables: train positive_distance nanmean: 0.789524 2019-10-09 02:33:19.521: INFO @log_variables: train negative_distance nanmean: 1.410622 2019-10-09 02:33:19.521: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:33:19.521: INFO @log_variables: valid loss mean: 0.447488 2019-10-09 02:33:19.521: INFO @log_variables: valid age_loss mean: 6.520245 2019-10-09 02:33:19.521: INFO @log_variables: valid gender_loss mean: 0.206327 2019-10-09 02:33:19.521: INFO @log_variables: valid matching_loss nanmean: 0.528860 2019-10-09 02:33:19.521: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:33:19.521: INFO @log_variables: valid age_mae mean: 7.003803 2019-10-09 02:33:19.521: INFO @log_variables: valid gender_accuracy mean: 0.922818 2019-10-09 02:33:19.521: INFO @log_variables: valid positive_distance nanmean: 0.781695 2019-10-09 02:33:19.521: INFO @log_variables: valid negative_distance nanmean: 1.363598 2019-10-09 02:33:19.521: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:33:19.521: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:33:21.975: INFO @metrics_hook: valid matching accuracy: 0.8839719373988127, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:33:22.387: INFO @decay_lr : LR updated to `0.00083908875` 2019-10-09 02:33:23.055: INFO @model : Quantizing and saving the model 2019-10-09 02:33:24.167: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:33:24.173: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:33:24.179: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:33:24.185: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:33:24.191: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:33:24.197: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:33:24.202: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:33:24.208: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:33:24.213: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:33:24.219: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:33:24.225: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:33:24.230: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:33:24.235: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:33:24.240: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:33:24.245: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:33:24.251: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:33:24.256: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:33:24.261: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:33:24.266: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:33:24.271: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:33:24.276: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:33:24.281: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:33:24.287: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:33:24.292: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:33:24.297: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 02:33:24.302: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 02:33:24.308: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 02:33:24.316: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 02:33:38.001: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 02:33:38.295: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 02:33:38.314: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 02:33:40.257: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 02:33:40.307: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 02:33:40.312: INFO @log_profile : T train: 178.275402 2019-10-09 02:33:40.313: INFO @log_profile : T valid: 8.282949 2019-10-09 02:33:40.314: INFO @log_profile : T read data: 0.983393 2019-10-09 02:33:40.314: INFO @log_profile : T hooks: 21.366843 2019-10-09 02:33:40.314: INFO @main_loop : Epoch 35 done 2019-10-09 02:33:40.314: INFO @main_loop : Training epoch 36 2019-10-09 02:36:49.147: INFO @log_variables: train loss mean: 0.350448 2019-10-09 02:36:49.147: INFO @log_variables: train age_loss mean: 5.776573 2019-10-09 02:36:49.147: INFO @log_variables: train gender_loss mean: 0.097137 2019-10-09 02:36:49.147: INFO @log_variables: train matching_loss nanmean: 0.411593 2019-10-09 02:36:49.147: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 02:36:49.148: INFO @log_variables: train age_mae mean: 6.258433 2019-10-09 02:36:49.148: INFO @log_variables: train gender_accuracy mean: 0.963225 2019-10-09 02:36:49.148: INFO @log_variables: train positive_distance nanmean: 0.787001 2019-10-09 02:36:49.148: INFO @log_variables: train negative_distance nanmean: 1.410430 2019-10-09 02:36:49.148: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:36:49.148: INFO @log_variables: valid loss mean: 0.459480 2019-10-09 02:36:49.148: INFO @log_variables: valid age_loss mean: 6.537497 2019-10-09 02:36:49.148: INFO @log_variables: valid gender_loss mean: 0.241567 2019-10-09 02:36:49.148: INFO @log_variables: valid matching_loss nanmean: 0.529072 2019-10-09 02:36:49.148: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:36:49.148: INFO @log_variables: valid age_mae mean: 7.021583 2019-10-09 02:36:49.148: INFO @log_variables: valid gender_accuracy mean: 0.913414 2019-10-09 02:36:49.148: INFO @log_variables: valid positive_distance nanmean: 0.783823 2019-10-09 02:36:49.148: INFO @log_variables: valid negative_distance nanmean: 1.363116 2019-10-09 02:36:49.148: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:36:49.148: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:36:51.334: INFO @metrics_hook: valid matching accuracy: 0.8834322719913653, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:36:51.795: INFO @decay_lr : LR updated to `0.0008348933` 2019-10-09 02:36:51.797: INFO @log_profile : T train: 178.443452 2019-10-09 02:36:51.797: INFO @log_profile : T valid: 8.332585 2019-10-09 02:36:51.797: INFO @log_profile : T read data: 1.413767 2019-10-09 02:36:51.797: INFO @log_profile : T hooks: 3.208942 2019-10-09 02:36:51.797: INFO @main_loop : Epoch 36 done 2019-10-09 02:36:51.797: INFO @main_loop : Training epoch 37 2019-10-09 02:40:00.516: INFO @log_variables: train loss mean: 0.348822 2019-10-09 02:40:00.516: INFO @log_variables: train age_loss mean: 5.786402 2019-10-09 02:40:00.516: INFO @log_variables: train gender_loss mean: 0.094560 2019-10-09 02:40:00.516: INFO @log_variables: train matching_loss nanmean: 0.408147 2019-10-09 02:40:00.516: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 02:40:00.517: INFO @log_variables: train age_mae mean: 6.267831 2019-10-09 02:40:00.517: INFO @log_variables: train gender_accuracy mean: 0.964217 2019-10-09 02:40:00.517: INFO @log_variables: train positive_distance nanmean: 0.784708 2019-10-09 02:40:00.517: INFO @log_variables: train negative_distance nanmean: 1.410016 2019-10-09 02:40:00.517: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:40:00.517: INFO @log_variables: valid loss mean: 0.456790 2019-10-09 02:40:00.517: INFO @log_variables: valid age_loss mean: 6.399135 2019-10-09 02:40:00.517: INFO @log_variables: valid gender_loss mean: 0.242961 2019-10-09 02:40:00.517: INFO @log_variables: valid matching_loss nanmean: 0.533176 2019-10-09 02:40:00.517: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:40:00.517: INFO @log_variables: valid age_mae mean: 6.882165 2019-10-09 02:40:00.517: INFO @log_variables: valid gender_accuracy mean: 0.911994 2019-10-09 02:40:00.517: INFO @log_variables: valid positive_distance nanmean: 0.784732 2019-10-09 02:40:00.517: INFO @log_variables: valid negative_distance nanmean: 1.356975 2019-10-09 02:40:00.517: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:40:00.517: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:40:02.717: INFO @metrics_hook: valid matching accuracy: 0.8831324578761168, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:40:03.166: INFO @decay_lr : LR updated to `0.00083071884` 2019-10-09 02:40:03.168: INFO @log_profile : T train: 178.327687 2019-10-09 02:40:03.168: INFO @log_profile : T valid: 8.329373 2019-10-09 02:40:03.168: INFO @log_profile : T read data: 1.413209 2019-10-09 02:40:03.168: INFO @log_profile : T hooks: 3.215373 2019-10-09 02:40:03.168: INFO @main_loop : Epoch 37 done 2019-10-09 02:40:03.168: INFO @main_loop : Training epoch 38 2019-10-09 02:43:12.164: INFO @log_variables: train loss mean: 0.346750 2019-10-09 02:43:12.165: INFO @log_variables: train age_loss mean: 5.758535 2019-10-09 02:43:12.165: INFO @log_variables: train gender_loss mean: 0.092668 2019-10-09 02:43:12.165: INFO @log_variables: train matching_loss nanmean: 0.406403 2019-10-09 02:43:12.165: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 02:43:12.165: INFO @log_variables: train age_mae mean: 6.239735 2019-10-09 02:43:12.165: INFO @log_variables: train gender_accuracy mean: 0.965125 2019-10-09 02:43:12.165: INFO @log_variables: train positive_distance nanmean: 0.784551 2019-10-09 02:43:12.165: INFO @log_variables: train negative_distance nanmean: 1.410029 2019-10-09 02:43:12.165: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:43:12.165: INFO @log_variables: valid loss mean: 0.466634 2019-10-09 02:43:12.165: INFO @log_variables: valid age_loss mean: 6.879697 2019-10-09 02:43:12.165: INFO @log_variables: valid gender_loss mean: 0.227818 2019-10-09 02:43:12.165: INFO @log_variables: valid matching_loss nanmean: 0.530779 2019-10-09 02:43:12.165: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:43:12.166: INFO @log_variables: valid age_mae mean: 7.362752 2019-10-09 02:43:12.166: INFO @log_variables: valid gender_accuracy mean: 0.916312 2019-10-09 02:43:12.166: INFO @log_variables: valid positive_distance nanmean: 0.788425 2019-10-09 02:43:12.166: INFO @log_variables: valid negative_distance nanmean: 1.359741 2019-10-09 02:43:12.166: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:43:12.166: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:43:14.606: INFO @metrics_hook: valid matching accuracy: 0.8867901900821491, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:43:15.030: INFO @decay_lr : LR updated to `0.00082656526` 2019-10-09 02:43:15.031: INFO @log_profile : T train: 178.984405 2019-10-09 02:43:15.031: INFO @log_profile : T valid: 8.374761 2019-10-09 02:43:15.031: INFO @log_profile : T read data: 1.000789 2019-10-09 02:43:15.031: INFO @log_profile : T hooks: 3.418622 2019-10-09 02:43:15.031: INFO @main_loop : Epoch 38 done 2019-10-09 02:43:15.031: INFO @main_loop : Training epoch 39 2019-10-09 02:46:24.555: INFO @log_variables: train loss mean: 0.346556 2019-10-09 02:46:24.555: INFO @log_variables: train age_loss mean: 5.752500 2019-10-09 02:46:24.556: INFO @log_variables: train gender_loss mean: 0.095920 2019-10-09 02:46:24.556: INFO @log_variables: train matching_loss nanmean: 0.403155 2019-10-09 02:46:24.556: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 02:46:24.556: INFO @log_variables: train age_mae mean: 6.233819 2019-10-09 02:46:24.556: INFO @log_variables: train gender_accuracy mean: 0.963997 2019-10-09 02:46:24.556: INFO @log_variables: train positive_distance nanmean: 0.782154 2019-10-09 02:46:24.556: INFO @log_variables: train negative_distance nanmean: 1.410044 2019-10-09 02:46:24.556: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:46:24.556: INFO @log_variables: valid loss mean: 0.448382 2019-10-09 02:46:24.556: INFO @log_variables: valid age_loss mean: 6.380239 2019-10-09 02:46:24.556: INFO @log_variables: valid gender_loss mean: 0.228375 2019-10-09 02:46:24.556: INFO @log_variables: valid matching_loss nanmean: 0.523585 2019-10-09 02:46:24.556: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:46:24.556: INFO @log_variables: valid age_mae mean: 6.863353 2019-10-09 02:46:24.556: INFO @log_variables: valid gender_accuracy mean: 0.916312 2019-10-09 02:46:24.556: INFO @log_variables: valid positive_distance nanmean: 0.773620 2019-10-09 02:46:24.556: INFO @log_variables: valid negative_distance nanmean: 1.360402 2019-10-09 02:46:24.557: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:46:24.557: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:46:26.749: INFO @metrics_hook: valid matching accuracy: 0.8857108592672543, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:46:27.184: INFO @decay_lr : LR updated to `0.00082243246` 2019-10-09 02:46:27.185: INFO @log_profile : T train: 179.123038 2019-10-09 02:46:27.185: INFO @log_profile : T valid: 8.334889 2019-10-09 02:46:27.186: INFO @log_profile : T read data: 1.434070 2019-10-09 02:46:27.186: INFO @log_profile : T hooks: 3.178572 2019-10-09 02:46:27.186: INFO @main_loop : Epoch 39 done 2019-10-09 02:46:27.186: INFO @main_loop : Training epoch 40 2019-10-09 02:49:35.883: INFO @log_variables: train loss mean: 0.343088 2019-10-09 02:49:35.883: INFO @log_variables: train age_loss mean: 5.693440 2019-10-09 02:49:35.883: INFO @log_variables: train gender_loss mean: 0.089775 2019-10-09 02:49:35.883: INFO @log_variables: train matching_loss nanmean: 0.404454 2019-10-09 02:49:35.883: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 02:49:35.883: INFO @log_variables: train age_mae mean: 6.174320 2019-10-09 02:49:35.883: INFO @log_variables: train gender_accuracy mean: 0.966073 2019-10-09 02:49:35.883: INFO @log_variables: train positive_distance nanmean: 0.783259 2019-10-09 02:49:35.883: INFO @log_variables: train negative_distance nanmean: 1.410046 2019-10-09 02:49:35.883: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:49:35.883: INFO @log_variables: valid loss mean: 0.466704 2019-10-09 02:49:35.883: INFO @log_variables: valid age_loss mean: 6.907081 2019-10-09 02:49:35.883: INFO @log_variables: valid gender_loss mean: 0.230661 2019-10-09 02:49:35.883: INFO @log_variables: valid matching_loss nanmean: 0.525413 2019-10-09 02:49:35.883: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:49:35.884: INFO @log_variables: valid age_mae mean: 7.390837 2019-10-09 02:49:35.884: INFO @log_variables: valid gender_accuracy mean: 0.917317 2019-10-09 02:49:35.884: INFO @log_variables: valid positive_distance nanmean: 0.776509 2019-10-09 02:49:35.884: INFO @log_variables: valid negative_distance nanmean: 1.360494 2019-10-09 02:49:35.884: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:49:35.884: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:49:38.187: INFO @metrics_hook: valid matching accuracy: 0.8833123463452659, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:49:38.659: INFO @decay_lr : LR updated to `0.0008183203` 2019-10-09 02:49:38.660: INFO @log_profile : T train: 178.708719 2019-10-09 02:49:38.660: INFO @log_profile : T valid: 8.317611 2019-10-09 02:49:38.660: INFO @log_profile : T read data: 1.008832 2019-10-09 02:49:38.660: INFO @log_profile : T hooks: 3.354257 2019-10-09 02:49:38.661: INFO @main_loop : Epoch 40 done 2019-10-09 02:49:38.661: INFO @main_loop : Training epoch 41 2019-10-09 02:52:47.794: INFO @log_variables: train loss mean: 0.341893 2019-10-09 02:52:47.795: INFO @log_variables: train age_loss mean: 5.673336 2019-10-09 02:52:47.795: INFO @log_variables: train gender_loss mean: 0.090316 2019-10-09 02:52:47.795: INFO @log_variables: train matching_loss nanmean: 0.402214 2019-10-09 02:52:47.795: INFO @log_variables: train is_face_loss mean: 0.000003 2019-10-09 02:52:47.795: INFO @log_variables: train age_mae mean: 6.154541 2019-10-09 02:52:47.795: INFO @log_variables: train gender_accuracy mean: 0.966107 2019-10-09 02:52:47.795: INFO @log_variables: train positive_distance nanmean: 0.782230 2019-10-09 02:52:47.795: INFO @log_variables: train negative_distance nanmean: 1.409991 2019-10-09 02:52:47.795: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:52:47.795: INFO @log_variables: valid loss mean: 0.458192 2019-10-09 02:52:47.795: INFO @log_variables: valid age_loss mean: 6.706943 2019-10-09 02:52:47.795: INFO @log_variables: valid gender_loss mean: 0.218027 2019-10-09 02:52:47.795: INFO @log_variables: valid matching_loss nanmean: 0.531672 2019-10-09 02:52:47.795: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:52:47.795: INFO @log_variables: valid age_mae mean: 7.191429 2019-10-09 02:52:47.795: INFO @log_variables: valid gender_accuracy mean: 0.920393 2019-10-09 02:52:47.795: INFO @log_variables: valid positive_distance nanmean: 0.779742 2019-10-09 02:52:47.796: INFO @log_variables: valid negative_distance nanmean: 1.362199 2019-10-09 02:52:47.796: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:52:47.796: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:52:49.843: INFO @metrics_hook: valid matching accuracy: 0.8844516399832104, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:52:50.261: INFO @decay_lr : LR updated to `0.0008142287` 2019-10-09 02:52:50.263: INFO @log_profile : T train: 178.717911 2019-10-09 02:52:50.263: INFO @log_profile : T valid: 8.335146 2019-10-09 02:52:50.263: INFO @log_profile : T read data: 1.419067 2019-10-09 02:52:50.263: INFO @log_profile : T hooks: 3.045605 2019-10-09 02:52:50.263: INFO @main_loop : Epoch 41 done 2019-10-09 02:52:50.263: INFO @main_loop : Training epoch 42 2019-10-09 02:55:59.069: INFO @log_variables: train loss mean: 0.336341 2019-10-09 02:55:59.069: INFO @log_variables: train age_loss mean: 5.595974 2019-10-09 02:55:59.069: INFO @log_variables: train gender_loss mean: 0.085084 2019-10-09 02:55:59.069: INFO @log_variables: train matching_loss nanmean: 0.397974 2019-10-09 02:55:59.069: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 02:55:59.069: INFO @log_variables: train age_mae mean: 6.076664 2019-10-09 02:55:59.069: INFO @log_variables: train gender_accuracy mean: 0.967961 2019-10-09 02:55:59.069: INFO @log_variables: train positive_distance nanmean: 0.780504 2019-10-09 02:55:59.070: INFO @log_variables: train negative_distance nanmean: 1.410280 2019-10-09 02:55:59.070: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:55:59.070: INFO @log_variables: valid loss mean: 0.460534 2019-10-09 02:55:59.070: INFO @log_variables: valid age_loss mean: 6.692597 2019-10-09 02:55:59.070: INFO @log_variables: valid gender_loss mean: 0.224273 2019-10-09 02:55:59.070: INFO @log_variables: valid matching_loss nanmean: 0.534123 2019-10-09 02:55:59.070: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:55:59.070: INFO @log_variables: valid age_mae mean: 7.175252 2019-10-09 02:55:59.070: INFO @log_variables: valid gender_accuracy mean: 0.916548 2019-10-09 02:55:59.070: INFO @log_variables: valid positive_distance nanmean: 0.776874 2019-10-09 02:55:59.070: INFO @log_variables: valid negative_distance nanmean: 1.358289 2019-10-09 02:55:59.070: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:55:59.070: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:56:01.151: INFO @metrics_hook: valid matching accuracy: 0.8836721232835641, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:56:01.570: INFO @decay_lr : LR updated to `0.00081015757` 2019-10-09 02:56:01.572: INFO @log_profile : T train: 178.372038 2019-10-09 02:56:01.572: INFO @log_profile : T valid: 8.348220 2019-10-09 02:56:01.572: INFO @log_profile : T read data: 1.436782 2019-10-09 02:56:01.572: INFO @log_profile : T hooks: 3.066218 2019-10-09 02:56:01.572: INFO @main_loop : Epoch 42 done 2019-10-09 02:56:01.572: INFO @main_loop : Training epoch 43 2019-10-09 02:59:10.260: INFO @log_variables: train loss mean: 0.336133 2019-10-09 02:59:10.260: INFO @log_variables: train age_loss mean: 5.588210 2019-10-09 02:59:10.260: INFO @log_variables: train gender_loss mean: 0.088151 2019-10-09 02:59:10.260: INFO @log_variables: train matching_loss nanmean: 0.395039 2019-10-09 02:59:10.260: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 02:59:10.260: INFO @log_variables: train age_mae mean: 6.068644 2019-10-09 02:59:10.260: INFO @log_variables: train gender_accuracy mean: 0.966951 2019-10-09 02:59:10.260: INFO @log_variables: train positive_distance nanmean: 0.778389 2019-10-09 02:59:10.260: INFO @log_variables: train negative_distance nanmean: 1.409936 2019-10-09 02:59:10.261: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 02:59:10.261: INFO @log_variables: valid loss mean: 0.453889 2019-10-09 02:59:10.261: INFO @log_variables: valid age_loss mean: 6.582947 2019-10-09 02:59:10.261: INFO @log_variables: valid gender_loss mean: 0.225107 2019-10-09 02:59:10.261: INFO @log_variables: valid matching_loss nanmean: 0.523656 2019-10-09 02:59:10.261: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 02:59:10.261: INFO @log_variables: valid age_mae mean: 7.065738 2019-10-09 02:59:10.261: INFO @log_variables: valid gender_accuracy mean: 0.915602 2019-10-09 02:59:10.261: INFO @log_variables: valid positive_distance nanmean: 0.783287 2019-10-09 02:59:10.261: INFO @log_variables: valid negative_distance nanmean: 1.364998 2019-10-09 02:59:10.261: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 02:59:10.261: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 02:59:12.859: INFO @metrics_hook: valid matching accuracy: 0.8861305990286023, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 02:59:13.265: INFO @decay_lr : LR updated to `0.0008061068` 2019-10-09 02:59:13.266: INFO @log_profile : T train: 178.654197 2019-10-09 02:59:13.267: INFO @log_profile : T valid: 8.339110 2019-10-09 02:59:13.267: INFO @log_profile : T read data: 1.024679 2019-10-09 02:59:13.267: INFO @log_profile : T hooks: 3.590576 2019-10-09 02:59:13.267: INFO @main_loop : Epoch 43 done 2019-10-09 02:59:13.267: INFO @main_loop : Training epoch 44 2019-10-09 03:02:22.075: INFO @log_variables: train loss mean: 0.334452 2019-10-09 03:02:22.075: INFO @log_variables: train age_loss mean: 5.590669 2019-10-09 03:02:22.075: INFO @log_variables: train gender_loss mean: 0.082042 2019-10-09 03:02:22.075: INFO @log_variables: train matching_loss nanmean: 0.395692 2019-10-09 03:02:22.075: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:02:22.075: INFO @log_variables: train age_mae mean: 6.071688 2019-10-09 03:02:22.075: INFO @log_variables: train gender_accuracy mean: 0.969149 2019-10-09 03:02:22.075: INFO @log_variables: train positive_distance nanmean: 0.779130 2019-10-09 03:02:22.075: INFO @log_variables: train negative_distance nanmean: 1.409922 2019-10-09 03:02:22.075: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:02:22.075: INFO @log_variables: valid loss mean: 0.449590 2019-10-09 03:02:22.075: INFO @log_variables: valid age_loss mean: 6.487594 2019-10-09 03:02:22.075: INFO @log_variables: valid gender_loss mean: 0.219266 2019-10-09 03:02:22.076: INFO @log_variables: valid matching_loss nanmean: 0.525702 2019-10-09 03:02:22.076: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:02:22.076: INFO @log_variables: valid age_mae mean: 6.970322 2019-10-09 03:02:22.076: INFO @log_variables: valid gender_accuracy mean: 0.921221 2019-10-09 03:02:22.076: INFO @log_variables: valid positive_distance nanmean: 0.782737 2019-10-09 03:02:22.076: INFO @log_variables: valid negative_distance nanmean: 1.361623 2019-10-09 03:02:22.076: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:02:22.076: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:02:24.502: INFO @metrics_hook: valid matching accuracy: 0.8853510823289561, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:02:24.939: INFO @decay_lr : LR updated to `0.00080207625` 2019-10-09 03:02:24.940: INFO @log_profile : T train: 178.355150 2019-10-09 03:02:24.940: INFO @log_profile : T valid: 8.300767 2019-10-09 03:02:24.940: INFO @log_profile : T read data: 1.492050 2019-10-09 03:02:24.940: INFO @log_profile : T hooks: 3.440404 2019-10-09 03:02:24.940: INFO @main_loop : Epoch 44 done 2019-10-09 03:02:24.940: INFO @main_loop : Training epoch 45 2019-10-09 03:05:33.859: INFO @log_variables: train loss mean: 0.334604 2019-10-09 03:05:33.860: INFO @log_variables: train age_loss mean: 5.564314 2019-10-09 03:05:33.860: INFO @log_variables: train gender_loss mean: 0.084030 2019-10-09 03:05:33.860: INFO @log_variables: train matching_loss nanmean: 0.396810 2019-10-09 03:05:33.860: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:05:33.860: INFO @log_variables: train age_mae mean: 6.044684 2019-10-09 03:05:33.860: INFO @log_variables: train gender_accuracy mean: 0.968605 2019-10-09 03:05:33.860: INFO @log_variables: train positive_distance nanmean: 0.779935 2019-10-09 03:05:33.860: INFO @log_variables: train negative_distance nanmean: 1.409775 2019-10-09 03:05:33.860: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:05:33.860: INFO @log_variables: valid loss mean: 0.459227 2019-10-09 03:05:33.860: INFO @log_variables: valid age_loss mean: 6.763336 2019-10-09 03:05:33.860: INFO @log_variables: valid gender_loss mean: 0.229091 2019-10-09 03:05:33.860: INFO @log_variables: valid matching_loss nanmean: 0.518178 2019-10-09 03:05:33.860: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:05:33.860: INFO @log_variables: valid age_mae mean: 7.247675 2019-10-09 03:05:33.860: INFO @log_variables: valid gender_accuracy mean: 0.918618 2019-10-09 03:05:33.860: INFO @log_variables: valid positive_distance nanmean: 0.780612 2019-10-09 03:05:33.860: INFO @log_variables: valid negative_distance nanmean: 1.365825 2019-10-09 03:05:33.861: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:05:33.861: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:05:36.232: INFO @metrics_hook: valid matching accuracy: 0.8858907477364034, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:05:36.642: INFO @decay_lr : LR updated to `0.00079806586` 2019-10-09 03:05:36.643: INFO @log_profile : T train: 178.521496 2019-10-09 03:05:36.643: INFO @log_profile : T valid: 8.338450 2019-10-09 03:05:36.643: INFO @log_profile : T read data: 1.403588 2019-10-09 03:05:36.643: INFO @log_profile : T hooks: 3.353359 2019-10-09 03:05:36.644: INFO @main_loop : Epoch 45 done 2019-10-09 03:05:36.644: INFO @main_loop : Training epoch 46 2019-10-09 03:08:45.536: INFO @log_variables: train loss mean: 0.332296 2019-10-09 03:08:45.536: INFO @log_variables: train age_loss mean: 5.526623 2019-10-09 03:08:45.536: INFO @log_variables: train gender_loss mean: 0.084022 2019-10-09 03:08:45.536: INFO @log_variables: train matching_loss nanmean: 0.393434 2019-10-09 03:08:45.536: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:08:45.536: INFO @log_variables: train age_mae mean: 6.007148 2019-10-09 03:08:45.536: INFO @log_variables: train gender_accuracy mean: 0.968107 2019-10-09 03:08:45.537: INFO @log_variables: train positive_distance nanmean: 0.779486 2019-10-09 03:08:45.537: INFO @log_variables: train negative_distance nanmean: 1.409565 2019-10-09 03:08:45.537: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:08:45.537: INFO @log_variables: valid loss mean: 0.463350 2019-10-09 03:08:45.537: INFO @log_variables: valid age_loss mean: 6.694499 2019-10-09 03:08:45.537: INFO @log_variables: valid gender_loss mean: 0.243211 2019-10-09 03:08:45.537: INFO @log_variables: valid matching_loss nanmean: 0.523725 2019-10-09 03:08:45.537: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:08:45.537: INFO @log_variables: valid age_mae mean: 7.176940 2019-10-09 03:08:45.537: INFO @log_variables: valid gender_accuracy mean: 0.912704 2019-10-09 03:08:45.537: INFO @log_variables: valid positive_distance nanmean: 0.778493 2019-10-09 03:08:45.537: INFO @log_variables: valid negative_distance nanmean: 1.362692 2019-10-09 03:08:45.537: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:08:45.537: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:08:47.795: INFO @metrics_hook: valid matching accuracy: 0.8887090004197398, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:08:48.227: INFO @decay_lr : LR updated to `0.00079407555` 2019-10-09 03:08:48.229: INFO @log_profile : T train: 178.864810 2019-10-09 03:08:48.229: INFO @log_profile : T valid: 8.350503 2019-10-09 03:08:48.229: INFO @log_profile : T read data: 1.006211 2019-10-09 03:08:48.229: INFO @log_profile : T hooks: 3.278151 2019-10-09 03:08:48.229: INFO @main_loop : Epoch 46 done 2019-10-09 03:08:48.229: INFO @main_loop : Training epoch 47 2019-10-09 03:11:56.780: INFO @log_variables: train loss mean: 0.328852 2019-10-09 03:11:56.781: INFO @log_variables: train age_loss mean: 5.471387 2019-10-09 03:11:56.781: INFO @log_variables: train gender_loss mean: 0.078615 2019-10-09 03:11:56.781: INFO @log_variables: train matching_loss nanmean: 0.393688 2019-10-09 03:11:56.781: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:11:56.781: INFO @log_variables: train age_mae mean: 5.951815 2019-10-09 03:11:56.781: INFO @log_variables: train gender_accuracy mean: 0.970595 2019-10-09 03:11:56.781: INFO @log_variables: train positive_distance nanmean: 0.777851 2019-10-09 03:11:56.781: INFO @log_variables: train negative_distance nanmean: 1.409670 2019-10-09 03:11:56.781: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:11:56.781: INFO @log_variables: valid loss mean: 0.461930 2019-10-09 03:11:56.781: INFO @log_variables: valid age_loss mean: 6.451803 2019-10-09 03:11:56.781: INFO @log_variables: valid gender_loss mean: 0.270394 2019-10-09 03:11:56.782: INFO @log_variables: valid matching_loss nanmean: 0.516407 2019-10-09 03:11:56.782: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:11:56.782: INFO @log_variables: valid age_mae mean: 6.934507 2019-10-09 03:11:56.782: INFO @log_variables: valid gender_accuracy mean: 0.908741 2019-10-09 03:11:56.782: INFO @log_variables: valid positive_distance nanmean: 0.782666 2019-10-09 03:11:56.782: INFO @log_variables: valid negative_distance nanmean: 1.367117 2019-10-09 03:11:56.782: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:11:56.782: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:11:58.788: INFO @metrics_hook: valid matching accuracy: 0.8874497811356958, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:11:59.205: INFO @decay_lr : LR updated to `0.00079010514` 2019-10-09 03:11:59.206: INFO @log_profile : T train: 178.187149 2019-10-09 03:11:59.207: INFO @log_profile : T valid: 8.302512 2019-10-09 03:11:59.207: INFO @log_profile : T read data: 1.396029 2019-10-09 03:11:59.207: INFO @log_profile : T hooks: 3.007130 2019-10-09 03:11:59.207: INFO @main_loop : Epoch 47 done 2019-10-09 03:11:59.207: INFO @main_loop : Training epoch 48 2019-10-09 03:15:08.432: INFO @log_variables: train loss mean: 0.326856 2019-10-09 03:15:08.433: INFO @log_variables: train age_loss mean: 5.446854 2019-10-09 03:15:08.433: INFO @log_variables: train gender_loss mean: 0.078949 2019-10-09 03:15:08.433: INFO @log_variables: train matching_loss nanmean: 0.389618 2019-10-09 03:15:08.433: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:15:08.433: INFO @log_variables: train age_mae mean: 5.927235 2019-10-09 03:15:08.433: INFO @log_variables: train gender_accuracy mean: 0.970386 2019-10-09 03:15:08.433: INFO @log_variables: train positive_distance nanmean: 0.777663 2019-10-09 03:15:08.433: INFO @log_variables: train negative_distance nanmean: 1.409760 2019-10-09 03:15:08.433: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:15:08.433: INFO @log_variables: valid loss mean: 0.466130 2019-10-09 03:15:08.433: INFO @log_variables: valid age_loss mean: 6.616357 2019-10-09 03:15:08.433: INFO @log_variables: valid gender_loss mean: 0.267068 2019-10-09 03:15:08.433: INFO @log_variables: valid matching_loss nanmean: 0.516300 2019-10-09 03:15:08.433: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:15:08.433: INFO @log_variables: valid age_mae mean: 7.099724 2019-10-09 03:15:08.433: INFO @log_variables: valid gender_accuracy mean: 0.913828 2019-10-09 03:15:08.434: INFO @log_variables: valid positive_distance nanmean: 0.786022 2019-10-09 03:15:08.434: INFO @log_variables: valid negative_distance nanmean: 1.372742 2019-10-09 03:15:08.434: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:15:08.434: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:15:10.595: INFO @metrics_hook: valid matching accuracy: 0.8864304131438508, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:15:11.030: INFO @decay_lr : LR updated to `0.0007861546` 2019-10-09 03:15:11.032: INFO @log_profile : T train: 179.222219 2019-10-09 03:15:11.032: INFO @log_profile : T valid: 8.324240 2019-10-09 03:15:11.032: INFO @log_profile : T read data: 1.034796 2019-10-09 03:15:11.032: INFO @log_profile : T hooks: 3.157766 2019-10-09 03:15:11.032: INFO @main_loop : Epoch 48 done 2019-10-09 03:15:11.032: INFO @main_loop : Training epoch 49 2019-10-09 03:18:20.063: INFO @log_variables: train loss mean: 0.326511 2019-10-09 03:18:20.063: INFO @log_variables: train age_loss mean: 5.456283 2019-10-09 03:18:20.063: INFO @log_variables: train gender_loss mean: 0.077333 2019-10-09 03:18:20.063: INFO @log_variables: train matching_loss nanmean: 0.389224 2019-10-09 03:18:20.063: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:18:20.063: INFO @log_variables: train age_mae mean: 5.936432 2019-10-09 03:18:20.063: INFO @log_variables: train gender_accuracy mean: 0.971455 2019-10-09 03:18:20.063: INFO @log_variables: train positive_distance nanmean: 0.776757 2019-10-09 03:18:20.063: INFO @log_variables: train negative_distance nanmean: 1.409630 2019-10-09 03:18:20.064: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:18:20.064: INFO @log_variables: valid loss mean: 0.444394 2019-10-09 03:18:20.064: INFO @log_variables: valid age_loss mean: 6.416575 2019-10-09 03:18:20.064: INFO @log_variables: valid gender_loss mean: 0.218382 2019-10-09 03:18:20.064: INFO @log_variables: valid matching_loss nanmean: 0.517581 2019-10-09 03:18:20.064: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:18:20.064: INFO @log_variables: valid age_mae mean: 6.898828 2019-10-09 03:18:20.064: INFO @log_variables: valid gender_accuracy mean: 0.920748 2019-10-09 03:18:20.064: INFO @log_variables: valid positive_distance nanmean: 0.789950 2019-10-09 03:18:20.064: INFO @log_variables: valid negative_distance nanmean: 1.371616 2019-10-09 03:18:20.064: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:18:20.064: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:18:22.158: INFO @metrics_hook: valid matching accuracy: 0.8844516399832104, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:18:22.574: INFO @decay_lr : LR updated to `0.0007822238` 2019-10-09 03:18:22.575: INFO @log_profile : T train: 178.601420 2019-10-09 03:18:22.575: INFO @log_profile : T valid: 8.299925 2019-10-09 03:18:22.575: INFO @log_profile : T read data: 1.451543 2019-10-09 03:18:22.575: INFO @log_profile : T hooks: 3.104508 2019-10-09 03:18:22.575: INFO @main_loop : Epoch 49 done 2019-10-09 03:18:22.575: INFO @main_loop : Training epoch 50 2019-10-09 03:21:31.443: INFO @log_variables: train loss mean: 0.323959 2019-10-09 03:21:31.443: INFO @log_variables: train age_loss mean: 5.392106 2019-10-09 03:21:31.443: INFO @log_variables: train gender_loss mean: 0.077693 2019-10-09 03:21:31.443: INFO @log_variables: train matching_loss nanmean: 0.387368 2019-10-09 03:21:31.443: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:21:31.443: INFO @log_variables: train age_mae mean: 5.872221 2019-10-09 03:21:31.443: INFO @log_variables: train gender_accuracy mean: 0.970876 2019-10-09 03:21:31.443: INFO @log_variables: train positive_distance nanmean: 0.776334 2019-10-09 03:21:31.443: INFO @log_variables: train negative_distance nanmean: 1.409834 2019-10-09 03:21:31.443: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:21:31.443: INFO @log_variables: valid loss mean: 0.448402 2019-10-09 03:21:31.443: INFO @log_variables: valid age_loss mean: 6.485366 2019-10-09 03:21:31.444: INFO @log_variables: valid gender_loss mean: 0.227868 2019-10-09 03:21:31.444: INFO @log_variables: valid matching_loss nanmean: 0.513642 2019-10-09 03:21:31.444: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:21:31.444: INFO @log_variables: valid age_mae mean: 6.968216 2019-10-09 03:21:31.444: INFO @log_variables: valid gender_accuracy mean: 0.919801 2019-10-09 03:21:31.444: INFO @log_variables: valid positive_distance nanmean: 0.787901 2019-10-09 03:21:31.444: INFO @log_variables: valid negative_distance nanmean: 1.369243 2019-10-09 03:21:31.444: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:21:31.444: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:21:33.630: INFO @metrics_hook: valid matching accuracy: 0.8851112310367573, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:21:34.072: INFO @decay_lr : LR updated to `0.0007783127` 2019-10-09 03:21:34.074: INFO @log_profile : T train: 178.434754 2019-10-09 03:21:34.074: INFO @log_profile : T valid: 8.324713 2019-10-09 03:21:34.074: INFO @log_profile : T read data: 1.445669 2019-10-09 03:21:34.074: INFO @log_profile : T hooks: 3.206837 2019-10-09 03:21:34.074: INFO @main_loop : Epoch 50 done 2019-10-09 03:21:34.074: INFO @main_loop : Training epoch 51 2019-10-09 03:24:42.448: INFO @log_variables: train loss mean: 0.321345 2019-10-09 03:24:42.448: INFO @log_variables: train age_loss mean: 5.337865 2019-10-09 03:24:42.448: INFO @log_variables: train gender_loss mean: 0.076406 2019-10-09 03:24:42.448: INFO @log_variables: train matching_loss nanmean: 0.385976 2019-10-09 03:24:42.448: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:24:42.448: INFO @log_variables: train age_mae mean: 5.817476 2019-10-09 03:24:42.448: INFO @log_variables: train gender_accuracy mean: 0.971623 2019-10-09 03:24:42.448: INFO @log_variables: train positive_distance nanmean: 0.774584 2019-10-09 03:24:42.449: INFO @log_variables: train negative_distance nanmean: 1.409668 2019-10-09 03:24:42.449: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:24:42.449: INFO @log_variables: valid loss mean: 0.457510 2019-10-09 03:24:42.449: INFO @log_variables: valid age_loss mean: 6.681996 2019-10-09 03:24:42.449: INFO @log_variables: valid gender_loss mean: 0.233691 2019-10-09 03:24:42.449: INFO @log_variables: valid matching_loss nanmean: 0.516389 2019-10-09 03:24:42.449: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:24:42.449: INFO @log_variables: valid age_mae mean: 7.164614 2019-10-09 03:24:42.449: INFO @log_variables: valid gender_accuracy mean: 0.923527 2019-10-09 03:24:42.449: INFO @log_variables: valid positive_distance nanmean: 0.784716 2019-10-09 03:24:42.449: INFO @log_variables: valid negative_distance nanmean: 1.370255 2019-10-09 03:24:42.449: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:24:42.449: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:24:45.105: INFO @metrics_hook: valid matching accuracy: 0.8856508964442046, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:24:45.549: INFO @decay_lr : LR updated to `0.0007744211` 2019-10-09 03:24:46.251: INFO @model : Quantizing and saving the model 2019-10-09 03:24:47.090: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 03:24:47.097: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 03:24:47.102: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 03:24:47.107: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 03:24:47.112: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 03:24:47.117: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 03:24:47.123: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 03:24:47.128: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 03:24:47.133: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 03:24:47.138: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 03:24:47.143: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 03:24:47.148: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 03:24:47.154: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 03:24:47.159: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 03:24:47.164: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 03:24:47.169: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 03:24:47.174: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 03:24:47.179: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 03:24:47.184: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 03:24:47.189: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 03:24:47.195: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 03:24:47.200: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 03:24:47.205: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 03:24:47.210: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 03:24:47.216: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 03:24:47.221: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 03:24:47.226: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 03:24:47.233: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 03:25:00.252: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 03:25:00.525: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 03:25:00.544: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 03:25:02.491: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 03:25:02.536: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 03:25:02.540: INFO @log_profile : T train: 178.194684 2019-10-09 03:25:02.541: INFO @log_profile : T valid: 8.504842 2019-10-09 03:25:02.541: INFO @log_profile : T read data: 1.034300 2019-10-09 03:25:02.541: INFO @log_profile : T hooks: 20.647560 2019-10-09 03:25:02.541: INFO @main_loop : Epoch 51 done 2019-10-09 03:25:02.541: INFO @main_loop : Training epoch 52 2019-10-09 03:28:11.230: INFO @log_variables: train loss mean: 0.321561 2019-10-09 03:28:11.230: INFO @log_variables: train age_loss mean: 5.367830 2019-10-09 03:28:11.230: INFO @log_variables: train gender_loss mean: 0.075182 2019-10-09 03:28:11.230: INFO @log_variables: train matching_loss nanmean: 0.384874 2019-10-09 03:28:11.230: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:28:11.230: INFO @log_variables: train age_mae mean: 5.847449 2019-10-09 03:28:11.231: INFO @log_variables: train gender_accuracy mean: 0.972194 2019-10-09 03:28:11.231: INFO @log_variables: train positive_distance nanmean: 0.774188 2019-10-09 03:28:11.231: INFO @log_variables: train negative_distance nanmean: 1.409446 2019-10-09 03:28:11.231: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:28:11.231: INFO @log_variables: valid loss mean: 0.484502 2019-10-09 03:28:11.231: INFO @log_variables: valid age_loss mean: 6.965912 2019-10-09 03:28:11.231: INFO @log_variables: valid gender_loss mean: 0.289274 2019-10-09 03:28:11.231: INFO @log_variables: valid matching_loss nanmean: 0.516092 2019-10-09 03:28:11.231: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:28:11.231: INFO @log_variables: valid age_mae mean: 7.450109 2019-10-09 03:28:11.231: INFO @log_variables: valid gender_accuracy mean: 0.903832 2019-10-09 03:28:11.231: INFO @log_variables: valid positive_distance nanmean: 0.783239 2019-10-09 03:28:11.231: INFO @log_variables: valid negative_distance nanmean: 1.368921 2019-10-09 03:28:11.231: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:28:11.231: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:28:13.524: INFO @metrics_hook: valid matching accuracy: 0.8838520117527133, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:28:13.951: INFO @decay_lr : LR updated to `0.000770549` 2019-10-09 03:28:13.953: INFO @log_profile : T train: 178.276477 2019-10-09 03:28:13.953: INFO @log_profile : T valid: 8.317477 2019-10-09 03:28:13.953: INFO @log_profile : T read data: 1.412067 2019-10-09 03:28:13.953: INFO @log_profile : T hooks: 3.319369 2019-10-09 03:28:13.953: INFO @main_loop : Epoch 52 done 2019-10-09 03:28:13.953: INFO @main_loop : Training epoch 53 2019-10-09 03:31:22.565: INFO @log_variables: train loss mean: 0.321007 2019-10-09 03:31:22.566: INFO @log_variables: train age_loss mean: 5.359701 2019-10-09 03:31:22.566: INFO @log_variables: train gender_loss mean: 0.074189 2019-10-09 03:31:22.566: INFO @log_variables: train matching_loss nanmean: 0.384962 2019-10-09 03:31:22.566: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:31:22.566: INFO @log_variables: train age_mae mean: 5.839510 2019-10-09 03:31:22.566: INFO @log_variables: train gender_accuracy mean: 0.972096 2019-10-09 03:31:22.566: INFO @log_variables: train positive_distance nanmean: 0.773653 2019-10-09 03:31:22.566: INFO @log_variables: train negative_distance nanmean: 1.409542 2019-10-09 03:31:22.566: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:31:22.566: INFO @log_variables: valid loss mean: 0.473025 2019-10-09 03:31:22.566: INFO @log_variables: valid age_loss mean: 6.713313 2019-10-09 03:31:22.566: INFO @log_variables: valid gender_loss mean: 0.278548 2019-10-09 03:31:22.566: INFO @log_variables: valid matching_loss nanmean: 0.516498 2019-10-09 03:31:22.566: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:31:22.566: INFO @log_variables: valid age_mae mean: 7.197127 2019-10-09 03:31:22.566: INFO @log_variables: valid gender_accuracy mean: 0.910516 2019-10-09 03:31:22.567: INFO @log_variables: valid positive_distance nanmean: 0.780202 2019-10-09 03:31:22.567: INFO @log_variables: valid negative_distance nanmean: 1.365439 2019-10-09 03:31:22.567: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:31:22.567: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:31:24.780: INFO @metrics_hook: valid matching accuracy: 0.8887689632427894, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:31:25.206: INFO @decay_lr : LR updated to `0.00076669623` 2019-10-09 03:31:25.207: INFO @log_profile : T train: 178.413951 2019-10-09 03:31:25.207: INFO @log_profile : T valid: 8.572650 2019-10-09 03:31:25.207: INFO @log_profile : T read data: 0.965677 2019-10-09 03:31:25.207: INFO @log_profile : T hooks: 3.216179 2019-10-09 03:31:25.207: INFO @main_loop : Epoch 53 done 2019-10-09 03:31:25.207: INFO @main_loop : Training epoch 54 2019-10-09 03:34:34.483: INFO @log_variables: train loss mean: 0.318808 2019-10-09 03:34:34.483: INFO @log_variables: train age_loss mean: 5.309837 2019-10-09 03:34:34.483: INFO @log_variables: train gender_loss mean: 0.074823 2019-10-09 03:34:34.483: INFO @log_variables: train matching_loss nanmean: 0.382497 2019-10-09 03:34:34.483: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:34:34.483: INFO @log_variables: train age_mae mean: 5.789687 2019-10-09 03:34:34.483: INFO @log_variables: train gender_accuracy mean: 0.972431 2019-10-09 03:34:34.483: INFO @log_variables: train positive_distance nanmean: 0.774271 2019-10-09 03:34:34.483: INFO @log_variables: train negative_distance nanmean: 1.409253 2019-10-09 03:34:34.483: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:34:34.484: INFO @log_variables: valid loss mean: 0.450509 2019-10-09 03:34:34.484: INFO @log_variables: valid age_loss mean: 6.565617 2019-10-09 03:34:34.484: INFO @log_variables: valid gender_loss mean: 0.225567 2019-10-09 03:34:34.484: INFO @log_variables: valid matching_loss nanmean: 0.514451 2019-10-09 03:34:34.484: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:34:34.484: INFO @log_variables: valid age_mae mean: 7.049520 2019-10-09 03:34:34.484: INFO @log_variables: valid gender_accuracy mean: 0.922049 2019-10-09 03:34:34.484: INFO @log_variables: valid positive_distance nanmean: 0.786043 2019-10-09 03:34:34.484: INFO @log_variables: valid negative_distance nanmean: 1.369047 2019-10-09 03:34:34.484: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:34:34.484: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:34:37.077: INFO @metrics_hook: valid matching accuracy: 0.8882292978353421, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:34:37.529: INFO @decay_lr : LR updated to `0.00076286274` 2019-10-09 03:34:37.530: INFO @log_profile : T train: 178.898621 2019-10-09 03:34:37.530: INFO @log_profile : T valid: 8.298088 2019-10-09 03:34:37.530: INFO @log_profile : T read data: 1.424209 2019-10-09 03:34:37.530: INFO @log_profile : T hooks: 3.617350 2019-10-09 03:34:37.530: INFO @main_loop : Epoch 54 done 2019-10-09 03:34:37.530: INFO @main_loop : Training epoch 55 2019-10-09 03:37:46.123: INFO @log_variables: train loss mean: 0.320557 2019-10-09 03:37:46.123: INFO @log_variables: train age_loss mean: 5.341995 2019-10-09 03:37:46.124: INFO @log_variables: train gender_loss mean: 0.075739 2019-10-09 03:37:46.124: INFO @log_variables: train matching_loss nanmean: 0.383789 2019-10-09 03:37:46.124: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:37:46.124: INFO @log_variables: train age_mae mean: 5.821511 2019-10-09 03:37:46.124: INFO @log_variables: train gender_accuracy mean: 0.972206 2019-10-09 03:37:46.124: INFO @log_variables: train positive_distance nanmean: 0.775024 2019-10-09 03:37:46.124: INFO @log_variables: train negative_distance nanmean: 1.408992 2019-10-09 03:37:46.124: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:37:46.124: INFO @log_variables: valid loss mean: 0.450950 2019-10-09 03:37:46.124: INFO @log_variables: valid age_loss mean: 6.519693 2019-10-09 03:37:46.124: INFO @log_variables: valid gender_loss mean: 0.234464 2019-10-09 03:37:46.124: INFO @log_variables: valid matching_loss nanmean: 0.511513 2019-10-09 03:37:46.124: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:37:46.124: INFO @log_variables: valid age_mae mean: 7.002954 2019-10-09 03:37:46.124: INFO @log_variables: valid gender_accuracy mean: 0.918796 2019-10-09 03:37:46.124: INFO @log_variables: valid positive_distance nanmean: 0.791120 2019-10-09 03:37:46.125: INFO @log_variables: valid negative_distance nanmean: 1.376291 2019-10-09 03:37:46.125: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:37:46.125: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:37:48.372: INFO @metrics_hook: valid matching accuracy: 0.8870900041973976, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:37:48.777: INFO @decay_lr : LR updated to `0.00075904845` 2019-10-09 03:37:48.779: INFO @log_profile : T train: 178.227822 2019-10-09 03:37:48.779: INFO @log_profile : T valid: 8.320724 2019-10-09 03:37:48.779: INFO @log_profile : T read data: 1.420440 2019-10-09 03:37:48.779: INFO @log_profile : T hooks: 3.194007 2019-10-09 03:37:48.779: INFO @main_loop : Epoch 55 done 2019-10-09 03:37:48.779: INFO @main_loop : Training epoch 56 2019-10-09 03:40:57.097: INFO @log_variables: train loss mean: 0.315443 2019-10-09 03:40:57.097: INFO @log_variables: train age_loss mean: 5.248359 2019-10-09 03:40:57.097: INFO @log_variables: train gender_loss mean: 0.073207 2019-10-09 03:40:57.097: INFO @log_variables: train matching_loss nanmean: 0.379831 2019-10-09 03:40:57.097: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:40:57.097: INFO @log_variables: train age_mae mean: 5.728230 2019-10-09 03:40:57.097: INFO @log_variables: train gender_accuracy mean: 0.973271 2019-10-09 03:40:57.097: INFO @log_variables: train positive_distance nanmean: 0.770227 2019-10-09 03:40:57.097: INFO @log_variables: train negative_distance nanmean: 1.409034 2019-10-09 03:40:57.097: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:40:57.097: INFO @log_variables: valid loss mean: 0.447867 2019-10-09 03:40:57.097: INFO @log_variables: valid age_loss mean: 6.448544 2019-10-09 03:40:57.098: INFO @log_variables: valid gender_loss mean: 0.231625 2019-10-09 03:40:57.098: INFO @log_variables: valid matching_loss nanmean: 0.511909 2019-10-09 03:40:57.098: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:40:57.098: INFO @log_variables: valid age_mae mean: 6.931377 2019-10-09 03:40:57.098: INFO @log_variables: valid gender_accuracy mean: 0.920215 2019-10-09 03:40:57.098: INFO @log_variables: valid positive_distance nanmean: 0.783588 2019-10-09 03:40:57.098: INFO @log_variables: valid negative_distance nanmean: 1.370264 2019-10-09 03:40:57.098: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:40:57.098: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:40:59.585: INFO @metrics_hook: valid matching accuracy: 0.8864304131438508, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:41:00.004: INFO @decay_lr : LR updated to `0.0007552532` 2019-10-09 03:41:00.005: INFO @log_profile : T train: 178.311077 2019-10-09 03:41:00.005: INFO @log_profile : T valid: 8.346379 2019-10-09 03:41:00.005: INFO @log_profile : T read data: 1.008037 2019-10-09 03:41:00.005: INFO @log_profile : T hooks: 3.474352 2019-10-09 03:41:00.005: INFO @main_loop : Epoch 56 done 2019-10-09 03:41:00.005: INFO @main_loop : Training epoch 57 2019-10-09 03:44:08.733: INFO @log_variables: train loss mean: 0.315011 2019-10-09 03:44:08.733: INFO @log_variables: train age_loss mean: 5.247668 2019-10-09 03:44:08.733: INFO @log_variables: train gender_loss mean: 0.072221 2019-10-09 03:44:08.733: INFO @log_variables: train matching_loss nanmean: 0.379545 2019-10-09 03:44:08.733: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:44:08.733: INFO @log_variables: train age_mae mean: 5.727240 2019-10-09 03:44:08.733: INFO @log_variables: train gender_accuracy mean: 0.973894 2019-10-09 03:44:08.733: INFO @log_variables: train positive_distance nanmean: 0.770716 2019-10-09 03:44:08.733: INFO @log_variables: train negative_distance nanmean: 1.408941 2019-10-09 03:44:08.733: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:44:08.734: INFO @log_variables: valid loss mean: 0.453263 2019-10-09 03:44:08.734: INFO @log_variables: valid age_loss mean: 6.420085 2019-10-09 03:44:08.734: INFO @log_variables: valid gender_loss mean: 0.245451 2019-10-09 03:44:08.734: INFO @log_variables: valid matching_loss nanmean: 0.517657 2019-10-09 03:44:08.734: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:44:08.734: INFO @log_variables: valid age_mae mean: 6.902515 2019-10-09 03:44:08.734: INFO @log_variables: valid gender_accuracy mean: 0.919920 2019-10-09 03:44:08.734: INFO @log_variables: valid positive_distance nanmean: 0.780516 2019-10-09 03:44:08.734: INFO @log_variables: valid negative_distance nanmean: 1.367840 2019-10-09 03:44:08.734: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:44:08.734: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:44:10.912: INFO @metrics_hook: valid matching accuracy: 0.8884091863044912, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:44:11.341: INFO @decay_lr : LR updated to `0.0007514769` 2019-10-09 03:44:11.343: INFO @log_profile : T train: 178.291786 2019-10-09 03:44:11.343: INFO @log_profile : T valid: 8.350313 2019-10-09 03:44:11.343: INFO @log_profile : T read data: 1.436972 2019-10-09 03:44:11.343: INFO @log_profile : T hooks: 3.170883 2019-10-09 03:44:11.343: INFO @main_loop : Epoch 57 done 2019-10-09 03:44:11.343: INFO @main_loop : Training epoch 58 2019-10-09 03:47:19.965: INFO @log_variables: train loss mean: 0.316217 2019-10-09 03:47:19.965: INFO @log_variables: train age_loss mean: 5.233219 2019-10-09 03:47:19.965: INFO @log_variables: train gender_loss mean: 0.076628 2019-10-09 03:47:19.965: INFO @log_variables: train matching_loss nanmean: 0.380321 2019-10-09 03:47:19.965: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:47:19.965: INFO @log_variables: train age_mae mean: 5.712450 2019-10-09 03:47:19.965: INFO @log_variables: train gender_accuracy mean: 0.971630 2019-10-09 03:47:19.965: INFO @log_variables: train positive_distance nanmean: 0.773454 2019-10-09 03:47:19.965: INFO @log_variables: train negative_distance nanmean: 1.408873 2019-10-09 03:47:19.965: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:47:19.965: INFO @log_variables: valid loss mean: 0.458299 2019-10-09 03:47:19.966: INFO @log_variables: valid age_loss mean: 6.797894 2019-10-09 03:47:19.966: INFO @log_variables: valid gender_loss mean: 0.220696 2019-10-09 03:47:19.966: INFO @log_variables: valid matching_loss nanmean: 0.520241 2019-10-09 03:47:19.966: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:47:19.966: INFO @log_variables: valid age_mae mean: 7.282135 2019-10-09 03:47:19.966: INFO @log_variables: valid gender_accuracy mean: 0.918678 2019-10-09 03:47:19.966: INFO @log_variables: valid positive_distance nanmean: 0.779134 2019-10-09 03:47:19.966: INFO @log_variables: valid negative_distance nanmean: 1.361793 2019-10-09 03:47:19.966: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:47:19.966: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:47:21.894: INFO @metrics_hook: valid matching accuracy: 0.8845715656293098, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:47:22.337: INFO @decay_lr : LR updated to `0.00074771955` 2019-10-09 03:47:22.338: INFO @log_profile : T train: 178.226703 2019-10-09 03:47:22.338: INFO @log_profile : T valid: 8.323838 2019-10-09 03:47:22.338: INFO @log_profile : T read data: 1.431292 2019-10-09 03:47:22.338: INFO @log_profile : T hooks: 2.927413 2019-10-09 03:47:22.338: INFO @main_loop : Epoch 58 done 2019-10-09 03:47:22.338: INFO @main_loop : Training epoch 59 2019-10-09 03:50:30.537: INFO @log_variables: train loss mean: 0.313009 2019-10-09 03:50:30.538: INFO @log_variables: train age_loss mean: 5.224754 2019-10-09 03:50:30.538: INFO @log_variables: train gender_loss mean: 0.070535 2019-10-09 03:50:30.538: INFO @log_variables: train matching_loss nanmean: 0.377316 2019-10-09 03:50:30.538: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:50:30.538: INFO @log_variables: train age_mae mean: 5.704381 2019-10-09 03:50:30.538: INFO @log_variables: train gender_accuracy mean: 0.973795 2019-10-09 03:50:30.538: INFO @log_variables: train positive_distance nanmean: 0.771688 2019-10-09 03:50:30.538: INFO @log_variables: train negative_distance nanmean: 1.408976 2019-10-09 03:50:30.538: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:50:30.538: INFO @log_variables: valid loss mean: 0.463808 2019-10-09 03:50:30.538: INFO @log_variables: valid age_loss mean: 6.929710 2019-10-09 03:50:30.538: INFO @log_variables: valid gender_loss mean: 0.233658 2019-10-09 03:50:30.538: INFO @log_variables: valid matching_loss nanmean: 0.511177 2019-10-09 03:50:30.538: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:50:30.538: INFO @log_variables: valid age_mae mean: 7.413945 2019-10-09 03:50:30.538: INFO @log_variables: valid gender_accuracy mean: 0.920925 2019-10-09 03:50:30.538: INFO @log_variables: valid positive_distance nanmean: 0.795418 2019-10-09 03:50:30.538: INFO @log_variables: valid negative_distance nanmean: 1.375097 2019-10-09 03:50:30.539: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:50:30.539: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:50:33.221: INFO @metrics_hook: valid matching accuracy: 0.8865503387899503, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:50:33.631: INFO @decay_lr : LR updated to `0.0007439809` 2019-10-09 03:50:33.632: INFO @log_profile : T train: 178.159206 2019-10-09 03:50:33.632: INFO @log_profile : T valid: 8.389925 2019-10-09 03:50:33.632: INFO @log_profile : T read data: 0.977180 2019-10-09 03:50:33.632: INFO @log_profile : T hooks: 3.680029 2019-10-09 03:50:33.632: INFO @main_loop : Epoch 59 done 2019-10-09 03:50:33.632: INFO @main_loop : Training epoch 60 2019-10-09 03:53:42.301: INFO @log_variables: train loss mean: 0.311069 2019-10-09 03:53:42.301: INFO @log_variables: train age_loss mean: 5.177659 2019-10-09 03:53:42.301: INFO @log_variables: train gender_loss mean: 0.072027 2019-10-09 03:53:42.301: INFO @log_variables: train matching_loss nanmean: 0.374520 2019-10-09 03:53:42.301: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:53:42.301: INFO @log_variables: train age_mae mean: 5.656698 2019-10-09 03:53:42.301: INFO @log_variables: train gender_accuracy mean: 0.973351 2019-10-09 03:53:42.301: INFO @log_variables: train positive_distance nanmean: 0.769162 2019-10-09 03:53:42.301: INFO @log_variables: train negative_distance nanmean: 1.409107 2019-10-09 03:53:42.301: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:53:42.301: INFO @log_variables: valid loss mean: 0.462888 2019-10-09 03:53:42.301: INFO @log_variables: valid age_loss mean: 6.337379 2019-10-09 03:53:42.302: INFO @log_variables: valid gender_loss mean: 0.285873 2019-10-09 03:53:42.302: INFO @log_variables: valid matching_loss nanmean: 0.515342 2019-10-09 03:53:42.302: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:53:42.302: INFO @log_variables: valid age_mae mean: 6.821020 2019-10-09 03:53:42.302: INFO @log_variables: valid gender_accuracy mean: 0.905429 2019-10-09 03:53:42.302: INFO @log_variables: valid positive_distance nanmean: 0.788039 2019-10-09 03:53:42.302: INFO @log_variables: valid negative_distance nanmean: 1.367075 2019-10-09 03:53:42.302: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:53:42.302: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:53:45.087: INFO @metrics_hook: valid matching accuracy: 0.8855909336211549, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:53:45.504: INFO @decay_lr : LR updated to `0.00074026105` 2019-10-09 03:53:45.506: INFO @log_profile : T train: 178.319964 2019-10-09 03:53:45.506: INFO @log_profile : T valid: 8.318367 2019-10-09 03:53:45.506: INFO @log_profile : T read data: 1.373475 2019-10-09 03:53:45.506: INFO @log_profile : T hooks: 3.777016 2019-10-09 03:53:45.506: INFO @main_loop : Epoch 60 done 2019-10-09 03:53:45.506: INFO @main_loop : Training epoch 61 2019-10-09 03:56:53.799: INFO @log_variables: train loss mean: 0.310366 2019-10-09 03:56:53.799: INFO @log_variables: train age_loss mean: 5.167452 2019-10-09 03:56:53.799: INFO @log_variables: train gender_loss mean: 0.069596 2019-10-09 03:56:53.799: INFO @log_variables: train matching_loss nanmean: 0.375794 2019-10-09 03:56:53.799: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 03:56:53.799: INFO @log_variables: train age_mae mean: 5.646400 2019-10-09 03:56:53.800: INFO @log_variables: train gender_accuracy mean: 0.974630 2019-10-09 03:56:53.800: INFO @log_variables: train positive_distance nanmean: 0.769933 2019-10-09 03:56:53.800: INFO @log_variables: train negative_distance nanmean: 1.408943 2019-10-09 03:56:53.800: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 03:56:53.800: INFO @log_variables: valid loss mean: 0.457633 2019-10-09 03:56:53.800: INFO @log_variables: valid age_loss mean: 6.862083 2019-10-09 03:56:53.800: INFO @log_variables: valid gender_loss mean: 0.222589 2019-10-09 03:56:53.800: INFO @log_variables: valid matching_loss nanmean: 0.509865 2019-10-09 03:56:53.800: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 03:56:53.800: INFO @log_variables: valid age_mae mean: 7.346130 2019-10-09 03:56:53.800: INFO @log_variables: valid gender_accuracy mean: 0.922226 2019-10-09 03:56:53.800: INFO @log_variables: valid positive_distance nanmean: 0.783861 2019-10-09 03:56:53.800: INFO @log_variables: valid negative_distance nanmean: 1.369930 2019-10-09 03:56:53.800: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 03:56:53.800: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 03:56:56.158: INFO @metrics_hook: valid matching accuracy: 0.8869700785512982, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 03:56:56.571: INFO @decay_lr : LR updated to `0.00073655974` 2019-10-09 03:56:56.572: INFO @log_profile : T train: 178.242324 2019-10-09 03:56:56.572: INFO @log_profile : T valid: 8.398070 2019-10-09 03:56:56.572: INFO @log_profile : T read data: 0.989058 2019-10-09 03:56:56.572: INFO @log_profile : T hooks: 3.350653 2019-10-09 03:56:56.572: INFO @main_loop : Epoch 61 done 2019-10-09 03:56:56.572: INFO @main_loop : Training epoch 62 2019-10-09 04:00:05.321: INFO @log_variables: train loss mean: 0.309045 2019-10-09 04:00:05.322: INFO @log_variables: train age_loss mean: 5.133913 2019-10-09 04:00:05.322: INFO @log_variables: train gender_loss mean: 0.067212 2019-10-09 04:00:05.322: INFO @log_variables: train matching_loss nanmean: 0.377436 2019-10-09 04:00:05.322: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:00:05.322: INFO @log_variables: train age_mae mean: 5.613149 2019-10-09 04:00:05.322: INFO @log_variables: train gender_accuracy mean: 0.975604 2019-10-09 04:00:05.322: INFO @log_variables: train positive_distance nanmean: 0.771439 2019-10-09 04:00:05.322: INFO @log_variables: train negative_distance nanmean: 1.409145 2019-10-09 04:00:05.322: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:00:05.322: INFO @log_variables: valid loss mean: 0.445511 2019-10-09 04:00:05.322: INFO @log_variables: valid age_loss mean: 6.459552 2019-10-09 04:00:05.322: INFO @log_variables: valid gender_loss mean: 0.219458 2019-10-09 04:00:05.323: INFO @log_variables: valid matching_loss nanmean: 0.515673 2019-10-09 04:00:05.323: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:00:05.323: INFO @log_variables: valid age_mae mean: 6.944054 2019-10-09 04:00:05.323: INFO @log_variables: valid gender_accuracy mean: 0.923232 2019-10-09 04:00:05.323: INFO @log_variables: valid positive_distance nanmean: 0.783995 2019-10-09 04:00:05.323: INFO @log_variables: valid negative_distance nanmean: 1.368106 2019-10-09 04:00:05.323: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:00:05.323: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:00:07.365: INFO @metrics_hook: valid matching accuracy: 0.885770822090304, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:00:07.784: INFO @decay_lr : LR updated to `0.00073287694` 2019-10-09 04:00:07.785: INFO @log_profile : T train: 178.368719 2019-10-09 04:00:07.785: INFO @log_profile : T valid: 8.332305 2019-10-09 04:00:07.785: INFO @log_profile : T read data: 1.388485 2019-10-09 04:00:07.785: INFO @log_profile : T hooks: 3.037483 2019-10-09 04:00:07.785: INFO @main_loop : Epoch 62 done 2019-10-09 04:00:07.785: INFO @main_loop : Training epoch 63 2019-10-09 04:03:16.563: INFO @log_variables: train loss mean: 0.307580 2019-10-09 04:03:16.563: INFO @log_variables: train age_loss mean: 5.135437 2019-10-09 04:03:16.563: INFO @log_variables: train gender_loss mean: 0.066523 2019-10-09 04:03:16.563: INFO @log_variables: train matching_loss nanmean: 0.373430 2019-10-09 04:03:16.563: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:03:16.563: INFO @log_variables: train age_mae mean: 5.614579 2019-10-09 04:03:16.563: INFO @log_variables: train gender_accuracy mean: 0.975671 2019-10-09 04:03:16.563: INFO @log_variables: train positive_distance nanmean: 0.767854 2019-10-09 04:03:16.563: INFO @log_variables: train negative_distance nanmean: 1.408990 2019-10-09 04:03:16.563: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:03:16.563: INFO @log_variables: valid loss mean: 0.454420 2019-10-09 04:03:16.563: INFO @log_variables: valid age_loss mean: 6.534575 2019-10-09 04:03:16.563: INFO @log_variables: valid gender_loss mean: 0.240270 2019-10-09 04:03:16.563: INFO @log_variables: valid matching_loss nanmean: 0.514973 2019-10-09 04:03:16.564: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:03:16.564: INFO @log_variables: valid age_mae mean: 7.018207 2019-10-09 04:03:16.564: INFO @log_variables: valid gender_accuracy mean: 0.917258 2019-10-09 04:03:16.564: INFO @log_variables: valid positive_distance nanmean: 0.782519 2019-10-09 04:03:16.564: INFO @log_variables: valid negative_distance nanmean: 1.367900 2019-10-09 04:03:16.564: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:03:16.564: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:03:18.707: INFO @metrics_hook: valid matching accuracy: 0.8889488517119386, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:03:19.117: INFO @decay_lr : LR updated to `0.00072921254` 2019-10-09 04:03:19.118: INFO @log_profile : T train: 178.356124 2019-10-09 04:03:19.119: INFO @log_profile : T valid: 8.343161 2019-10-09 04:03:19.119: INFO @log_profile : T read data: 1.424647 2019-10-09 04:03:19.119: INFO @log_profile : T hooks: 3.125006 2019-10-09 04:03:19.119: INFO @main_loop : Epoch 63 done 2019-10-09 04:03:19.119: INFO @main_loop : Training epoch 64 2019-10-09 04:06:27.469: INFO @log_variables: train loss mean: 0.308582 2019-10-09 04:06:27.469: INFO @log_variables: train age_loss mean: 5.162192 2019-10-09 04:06:27.469: INFO @log_variables: train gender_loss mean: 0.068620 2019-10-09 04:06:27.469: INFO @log_variables: train matching_loss nanmean: 0.371765 2019-10-09 04:06:27.469: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:06:27.469: INFO @log_variables: train age_mae mean: 5.641142 2019-10-09 04:06:27.469: INFO @log_variables: train gender_accuracy mean: 0.974747 2019-10-09 04:06:27.469: INFO @log_variables: train positive_distance nanmean: 0.768887 2019-10-09 04:06:27.469: INFO @log_variables: train negative_distance nanmean: 1.409098 2019-10-09 04:06:27.470: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:06:27.470: INFO @log_variables: valid loss mean: 0.455443 2019-10-09 04:06:27.470: INFO @log_variables: valid age_loss mean: 6.635876 2019-10-09 04:06:27.470: INFO @log_variables: valid gender_loss mean: 0.234659 2019-10-09 04:06:27.470: INFO @log_variables: valid matching_loss nanmean: 0.513626 2019-10-09 04:06:27.470: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:06:27.470: INFO @log_variables: valid age_mae mean: 7.119685 2019-10-09 04:06:27.470: INFO @log_variables: valid gender_accuracy mean: 0.916726 2019-10-09 04:06:27.470: INFO @log_variables: valid positive_distance nanmean: 0.784415 2019-10-09 04:06:27.470: INFO @log_variables: valid negative_distance nanmean: 1.368642 2019-10-09 04:06:27.470: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:06:27.470: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:06:29.827: INFO @metrics_hook: valid matching accuracy: 0.886190561851652, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:06:30.232: INFO @decay_lr : LR updated to `0.00072556647` 2019-10-09 04:06:30.234: INFO @log_profile : T train: 178.386611 2019-10-09 04:06:30.234: INFO @log_profile : T valid: 8.327973 2019-10-09 04:06:30.234: INFO @log_profile : T read data: 0.988318 2019-10-09 04:06:30.234: INFO @log_profile : T hooks: 3.327414 2019-10-09 04:06:30.234: INFO @main_loop : Epoch 64 done 2019-10-09 04:06:30.234: INFO @main_loop : Training epoch 65 2019-10-09 04:09:39.114: INFO @log_variables: train loss mean: 0.308429 2019-10-09 04:09:39.114: INFO @log_variables: train age_loss mean: 5.155739 2019-10-09 04:09:39.114: INFO @log_variables: train gender_loss mean: 0.067477 2019-10-09 04:09:39.114: INFO @log_variables: train matching_loss nanmean: 0.373080 2019-10-09 04:09:39.114: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:09:39.114: INFO @log_variables: train age_mae mean: 5.634716 2019-10-09 04:09:39.114: INFO @log_variables: train gender_accuracy mean: 0.974566 2019-10-09 04:09:39.114: INFO @log_variables: train positive_distance nanmean: 0.769433 2019-10-09 04:09:39.114: INFO @log_variables: train negative_distance nanmean: 1.409005 2019-10-09 04:09:39.114: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:09:39.114: INFO @log_variables: valid loss mean: 0.453817 2019-10-09 04:09:39.114: INFO @log_variables: valid age_loss mean: 6.547652 2019-10-09 04:09:39.114: INFO @log_variables: valid gender_loss mean: 0.240545 2019-10-09 04:09:39.114: INFO @log_variables: valid matching_loss nanmean: 0.511522 2019-10-09 04:09:39.115: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:09:39.115: INFO @log_variables: valid age_mae mean: 7.031164 2019-10-09 04:09:39.115: INFO @log_variables: valid gender_accuracy mean: 0.920393 2019-10-09 04:09:39.115: INFO @log_variables: valid positive_distance nanmean: 0.788828 2019-10-09 04:09:39.115: INFO @log_variables: valid negative_distance nanmean: 1.370875 2019-10-09 04:09:39.115: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:09:39.115: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:09:41.474: INFO @metrics_hook: valid matching accuracy: 0.8867302272590993, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:09:41.902: INFO @decay_lr : LR updated to `0.0007219386` 2019-10-09 04:09:41.903: INFO @log_profile : T train: 178.384004 2019-10-09 04:09:41.903: INFO @log_profile : T valid: 8.426426 2019-10-09 04:09:41.903: INFO @log_profile : T read data: 1.410997 2019-10-09 04:09:41.903: INFO @log_profile : T hooks: 3.364027 2019-10-09 04:09:41.903: INFO @main_loop : Epoch 65 done 2019-10-09 04:09:41.903: INFO @main_loop : Training epoch 66 2019-10-09 04:12:50.596: INFO @log_variables: train loss mean: 0.305581 2019-10-09 04:12:50.596: INFO @log_variables: train age_loss mean: 5.100749 2019-10-09 04:12:50.596: INFO @log_variables: train gender_loss mean: 0.064120 2019-10-09 04:12:50.596: INFO @log_variables: train matching_loss nanmean: 0.373105 2019-10-09 04:12:50.596: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:12:50.596: INFO @log_variables: train age_mae mean: 5.579633 2019-10-09 04:12:50.596: INFO @log_variables: train gender_accuracy mean: 0.975922 2019-10-09 04:12:50.596: INFO @log_variables: train positive_distance nanmean: 0.768915 2019-10-09 04:12:50.596: INFO @log_variables: train negative_distance nanmean: 1.409016 2019-10-09 04:12:50.596: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:12:50.597: INFO @log_variables: valid loss mean: 0.453229 2019-10-09 04:12:50.597: INFO @log_variables: valid age_loss mean: 6.588512 2019-10-09 04:12:50.597: INFO @log_variables: valid gender_loss mean: 0.233440 2019-10-09 04:12:50.597: INFO @log_variables: valid matching_loss nanmean: 0.512717 2019-10-09 04:12:50.597: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:12:50.597: INFO @log_variables: valid age_mae mean: 7.072154 2019-10-09 04:12:50.597: INFO @log_variables: valid gender_accuracy mean: 0.919151 2019-10-09 04:12:50.597: INFO @log_variables: valid positive_distance nanmean: 0.787208 2019-10-09 04:12:50.597: INFO @log_variables: valid negative_distance nanmean: 1.368550 2019-10-09 04:12:50.597: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:12:50.597: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:12:53.005: INFO @metrics_hook: valid matching accuracy: 0.8887090004197398, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:12:53.432: INFO @decay_lr : LR updated to `0.0007183289` 2019-10-09 04:12:53.433: INFO @log_profile : T train: 178.330148 2019-10-09 04:12:53.433: INFO @log_profile : T valid: 8.332591 2019-10-09 04:12:53.433: INFO @log_profile : T read data: 1.372926 2019-10-09 04:12:53.434: INFO @log_profile : T hooks: 3.409525 2019-10-09 04:12:53.434: INFO @main_loop : Epoch 66 done 2019-10-09 04:12:53.434: INFO @main_loop : Training epoch 67 2019-10-09 04:16:01.515: INFO @log_variables: train loss mean: 0.303879 2019-10-09 04:16:01.516: INFO @log_variables: train age_loss mean: 5.074842 2019-10-09 04:16:01.516: INFO @log_variables: train gender_loss mean: 0.065698 2019-10-09 04:16:01.516: INFO @log_variables: train matching_loss nanmean: 0.368843 2019-10-09 04:16:01.516: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:16:01.516: INFO @log_variables: train age_mae mean: 5.553988 2019-10-09 04:16:01.516: INFO @log_variables: train gender_accuracy mean: 0.975842 2019-10-09 04:16:01.517: INFO @log_variables: train positive_distance nanmean: 0.767058 2019-10-09 04:16:01.517: INFO @log_variables: train negative_distance nanmean: 1.409023 2019-10-09 04:16:01.517: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:16:01.517: INFO @log_variables: valid loss mean: 0.450251 2019-10-09 04:16:01.517: INFO @log_variables: valid age_loss mean: 6.441759 2019-10-09 04:16:01.517: INFO @log_variables: valid gender_loss mean: 0.242420 2019-10-09 04:16:01.518: INFO @log_variables: valid matching_loss nanmean: 0.509183 2019-10-09 04:16:01.518: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:16:01.518: INFO @log_variables: valid age_mae mean: 6.924718 2019-10-09 04:16:01.518: INFO @log_variables: valid gender_accuracy mean: 0.919387 2019-10-09 04:16:01.518: INFO @log_variables: valid positive_distance nanmean: 0.795830 2019-10-09 04:16:01.518: INFO @log_variables: valid negative_distance nanmean: 1.374021 2019-10-09 04:16:01.519: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:16:01.519: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:16:04.799: INFO @metrics_hook: valid matching accuracy: 0.8866702644360497, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:16:05.215: INFO @decay_lr : LR updated to `0.0007147373` 2019-10-09 04:16:05.217: INFO @log_profile : T train: 178.123836 2019-10-09 04:16:05.217: INFO @log_profile : T valid: 8.257386 2019-10-09 04:16:05.217: INFO @log_profile : T read data: 0.992601 2019-10-09 04:16:05.217: INFO @log_profile : T hooks: 4.323540 2019-10-09 04:16:05.217: INFO @main_loop : Epoch 67 done 2019-10-09 04:16:05.217: INFO @main_loop : Training epoch 68 2019-10-09 04:19:13.728: INFO @log_variables: train loss mean: 0.301807 2019-10-09 04:19:13.728: INFO @log_variables: train age_loss mean: 5.018670 2019-10-09 04:19:13.728: INFO @log_variables: train gender_loss mean: 0.062911 2019-10-09 04:19:13.728: INFO @log_variables: train matching_loss nanmean: 0.370824 2019-10-09 04:19:13.728: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:19:13.728: INFO @log_variables: train age_mae mean: 5.497475 2019-10-09 04:19:13.728: INFO @log_variables: train gender_accuracy mean: 0.976952 2019-10-09 04:19:13.728: INFO @log_variables: train positive_distance nanmean: 0.767373 2019-10-09 04:19:13.728: INFO @log_variables: train negative_distance nanmean: 1.409348 2019-10-09 04:19:13.728: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:19:13.728: INFO @log_variables: valid loss mean: 0.451239 2019-10-09 04:19:13.728: INFO @log_variables: valid age_loss mean: 6.413338 2019-10-09 04:19:13.729: INFO @log_variables: valid gender_loss mean: 0.248852 2019-10-09 04:19:13.729: INFO @log_variables: valid matching_loss nanmean: 0.508654 2019-10-09 04:19:13.729: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:19:13.729: INFO @log_variables: valid age_mae mean: 6.896028 2019-10-09 04:19:13.729: INFO @log_variables: valid gender_accuracy mean: 0.917968 2019-10-09 04:19:13.729: INFO @log_variables: valid positive_distance nanmean: 0.790288 2019-10-09 04:19:13.729: INFO @log_variables: valid negative_distance nanmean: 1.371969 2019-10-09 04:19:13.729: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:19:13.729: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:19:16.109: INFO @metrics_hook: valid matching accuracy: 0.8874497811356958, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:19:16.535: INFO @decay_lr : LR updated to `0.0007111636` 2019-10-09 04:19:16.536: INFO @log_profile : T train: 178.015169 2019-10-09 04:19:16.537: INFO @log_profile : T valid: 8.387595 2019-10-09 04:19:16.537: INFO @log_profile : T read data: 1.454139 2019-10-09 04:19:16.537: INFO @log_profile : T hooks: 3.375469 2019-10-09 04:19:16.537: INFO @main_loop : Epoch 68 done 2019-10-09 04:19:16.537: INFO @main_loop : Training epoch 69 2019-10-09 04:22:24.691: INFO @log_variables: train loss mean: 0.303231 2019-10-09 04:22:24.691: INFO @log_variables: train age_loss mean: 5.062128 2019-10-09 04:22:24.691: INFO @log_variables: train gender_loss mean: 0.064334 2019-10-09 04:22:24.691: INFO @log_variables: train matching_loss nanmean: 0.369470 2019-10-09 04:22:24.691: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:22:24.691: INFO @log_variables: train age_mae mean: 5.540844 2019-10-09 04:22:24.691: INFO @log_variables: train gender_accuracy mean: 0.976744 2019-10-09 04:22:24.691: INFO @log_variables: train positive_distance nanmean: 0.767097 2019-10-09 04:22:24.691: INFO @log_variables: train negative_distance nanmean: 1.409062 2019-10-09 04:22:24.691: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:22:24.691: INFO @log_variables: valid loss mean: 0.457179 2019-10-09 04:22:24.691: INFO @log_variables: valid age_loss mean: 6.561749 2019-10-09 04:22:24.691: INFO @log_variables: valid gender_loss mean: 0.250831 2019-10-09 04:22:24.691: INFO @log_variables: valid matching_loss nanmean: 0.510248 2019-10-09 04:22:24.692: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:22:24.692: INFO @log_variables: valid age_mae mean: 7.046217 2019-10-09 04:22:24.692: INFO @log_variables: valid gender_accuracy mean: 0.918678 2019-10-09 04:22:24.692: INFO @log_variables: valid positive_distance nanmean: 0.779623 2019-10-09 04:22:24.692: INFO @log_variables: valid negative_distance nanmean: 1.366615 2019-10-09 04:22:24.692: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:22:24.692: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:22:26.689: INFO @metrics_hook: valid matching accuracy: 0.8860706362055526, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:22:27.111: INFO @decay_lr : LR updated to `0.00070760783` 2019-10-09 04:22:27.112: INFO @log_profile : T train: 178.204197 2019-10-09 04:22:27.112: INFO @log_profile : T valid: 8.327577 2019-10-09 04:22:27.112: INFO @log_profile : T read data: 0.969100 2019-10-09 04:22:27.112: INFO @log_profile : T hooks: 2.990188 2019-10-09 04:22:27.112: INFO @main_loop : Epoch 69 done 2019-10-09 04:22:27.112: INFO @main_loop : Training epoch 70 2019-10-09 04:25:35.752: INFO @log_variables: train loss mean: 0.298386 2019-10-09 04:25:35.752: INFO @log_variables: train age_loss mean: 4.989705 2019-10-09 04:25:35.752: INFO @log_variables: train gender_loss mean: 0.060742 2019-10-09 04:25:35.752: INFO @log_variables: train matching_loss nanmean: 0.365284 2019-10-09 04:25:35.753: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:25:35.753: INFO @log_variables: train age_mae mean: 5.468202 2019-10-09 04:25:35.753: INFO @log_variables: train gender_accuracy mean: 0.978182 2019-10-09 04:25:35.753: INFO @log_variables: train positive_distance nanmean: 0.765888 2019-10-09 04:25:35.753: INFO @log_variables: train negative_distance nanmean: 1.409285 2019-10-09 04:25:35.753: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:25:35.753: INFO @log_variables: valid loss mean: 0.455838 2019-10-09 04:25:35.753: INFO @log_variables: valid age_loss mean: 6.545679 2019-10-09 04:25:35.753: INFO @log_variables: valid gender_loss mean: 0.245257 2019-10-09 04:25:35.753: INFO @log_variables: valid matching_loss nanmean: 0.513272 2019-10-09 04:25:35.753: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:25:35.753: INFO @log_variables: valid age_mae mean: 7.029262 2019-10-09 04:25:35.753: INFO @log_variables: valid gender_accuracy mean: 0.917495 2019-10-09 04:25:35.753: INFO @log_variables: valid positive_distance nanmean: 0.776141 2019-10-09 04:25:35.753: INFO @log_variables: valid negative_distance nanmean: 1.366120 2019-10-09 04:25:35.753: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:25:35.753: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:25:37.995: INFO @metrics_hook: valid matching accuracy: 0.8881093721892427, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:25:38.416: INFO @decay_lr : LR updated to `0.0007040698` 2019-10-09 04:25:38.417: INFO @log_profile : T train: 178.264647 2019-10-09 04:25:38.417: INFO @log_profile : T valid: 8.343375 2019-10-09 04:25:38.417: INFO @log_profile : T read data: 1.387543 2019-10-09 04:25:38.418: INFO @log_profile : T hooks: 3.223618 2019-10-09 04:25:38.418: INFO @main_loop : Epoch 70 done 2019-10-09 04:25:38.418: INFO @main_loop : Training epoch 71 2019-10-09 04:28:47.019: INFO @log_variables: train loss mean: 0.300961 2019-10-09 04:28:47.019: INFO @log_variables: train age_loss mean: 5.009113 2019-10-09 04:28:47.019: INFO @log_variables: train gender_loss mean: 0.062802 2019-10-09 04:28:47.019: INFO @log_variables: train matching_loss nanmean: 0.369265 2019-10-09 04:28:47.019: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:28:47.019: INFO @log_variables: train age_mae mean: 5.487596 2019-10-09 04:28:47.019: INFO @log_variables: train gender_accuracy mean: 0.977025 2019-10-09 04:28:47.020: INFO @log_variables: train positive_distance nanmean: 0.766957 2019-10-09 04:28:47.020: INFO @log_variables: train negative_distance nanmean: 1.409005 2019-10-09 04:28:47.020: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:28:47.020: INFO @log_variables: valid loss mean: 0.451630 2019-10-09 04:28:47.020: INFO @log_variables: valid age_loss mean: 6.398421 2019-10-09 04:28:47.020: INFO @log_variables: valid gender_loss mean: 0.249658 2019-10-09 04:28:47.020: INFO @log_variables: valid matching_loss nanmean: 0.510553 2019-10-09 04:28:47.020: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:28:47.020: INFO @log_variables: valid age_mae mean: 6.881367 2019-10-09 04:28:47.020: INFO @log_variables: valid gender_accuracy mean: 0.919446 2019-10-09 04:28:47.020: INFO @log_variables: valid positive_distance nanmean: 0.784355 2019-10-09 04:28:47.020: INFO @log_variables: valid negative_distance nanmean: 1.371461 2019-10-09 04:28:47.020: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:28:47.020: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:28:49.084: INFO @metrics_hook: valid matching accuracy: 0.88864903759669, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:28:49.498: INFO @decay_lr : LR updated to `0.00070054946` 2019-10-09 04:28:49.499: INFO @log_profile : T train: 178.247371 2019-10-09 04:28:49.499: INFO @log_profile : T valid: 8.331576 2019-10-09 04:28:49.499: INFO @log_profile : T read data: 1.351919 2019-10-09 04:28:49.499: INFO @log_profile : T hooks: 3.063678 2019-10-09 04:28:49.499: INFO @main_loop : Epoch 71 done 2019-10-09 04:28:49.499: INFO @main_loop : Training epoch 72 2019-10-09 04:31:57.769: INFO @log_variables: train loss mean: 0.299216 2019-10-09 04:31:57.769: INFO @log_variables: train age_loss mean: 5.005749 2019-10-09 04:31:57.769: INFO @log_variables: train gender_loss mean: 0.061017 2019-10-09 04:31:57.769: INFO @log_variables: train matching_loss nanmean: 0.365979 2019-10-09 04:31:57.769: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:31:57.769: INFO @log_variables: train age_mae mean: 5.484605 2019-10-09 04:31:57.769: INFO @log_variables: train gender_accuracy mean: 0.977760 2019-10-09 04:31:57.769: INFO @log_variables: train positive_distance nanmean: 0.765720 2019-10-09 04:31:57.769: INFO @log_variables: train negative_distance nanmean: 1.408908 2019-10-09 04:31:57.769: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:31:57.769: INFO @log_variables: valid loss mean: 0.451570 2019-10-09 04:31:57.769: INFO @log_variables: valid age_loss mean: 6.666439 2019-10-09 04:31:57.769: INFO @log_variables: valid gender_loss mean: 0.225197 2019-10-09 04:31:57.769: INFO @log_variables: valid matching_loss nanmean: 0.508026 2019-10-09 04:31:57.769: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:31:57.769: INFO @log_variables: valid age_mae mean: 7.150404 2019-10-09 04:31:57.769: INFO @log_variables: valid gender_accuracy mean: 0.925242 2019-10-09 04:31:57.770: INFO @log_variables: valid positive_distance nanmean: 0.781278 2019-10-09 04:31:57.770: INFO @log_variables: valid negative_distance nanmean: 1.369795 2019-10-09 04:31:57.770: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:31:57.770: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:32:00.034: INFO @metrics_hook: valid matching accuracy: 0.887209929843497, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:32:00.482: INFO @decay_lr : LR updated to `0.0006970467` 2019-10-09 04:32:01.184: INFO @model : Quantizing and saving the model 2019-10-09 04:32:02.310: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 04:32:02.316: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 04:32:02.322: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 04:32:02.328: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 04:32:02.334: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 04:32:02.339: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 04:32:02.345: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 04:32:02.350: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 04:32:02.355: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 04:32:02.361: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 04:32:02.368: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 04:32:02.374: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 04:32:02.379: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 04:32:02.384: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 04:32:02.390: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 04:32:02.395: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 04:32:02.401: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 04:32:02.406: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 04:32:02.411: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 04:32:02.417: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 04:32:02.422: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 04:32:02.427: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 04:32:02.433: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 04:32:02.438: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 04:32:02.444: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 04:32:02.449: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 04:32:02.454: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 04:32:02.462: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 04:32:16.374: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 04:32:16.672: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 04:32:16.690: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 04:32:18.631: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 04:32:18.674: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 04:32:18.677: INFO @log_profile : T train: 178.283252 2019-10-09 04:32:18.678: INFO @log_profile : T valid: 8.334814 2019-10-09 04:32:18.680: INFO @log_profile : T read data: 0.990415 2019-10-09 04:32:18.680: INFO @log_profile : T hooks: 21.482703 2019-10-09 04:32:18.681: INFO @main_loop : Epoch 72 done 2019-10-09 04:32:18.681: INFO @main_loop : Training epoch 73 2019-10-09 04:35:27.277: INFO @log_variables: train loss mean: 0.298997 2019-10-09 04:35:27.277: INFO @log_variables: train age_loss mean: 4.991519 2019-10-09 04:35:27.277: INFO @log_variables: train gender_loss mean: 0.062178 2019-10-09 04:35:27.278: INFO @log_variables: train matching_loss nanmean: 0.365560 2019-10-09 04:35:27.278: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:35:27.278: INFO @log_variables: train age_mae mean: 5.470082 2019-10-09 04:35:27.278: INFO @log_variables: train gender_accuracy mean: 0.977080 2019-10-09 04:35:27.278: INFO @log_variables: train positive_distance nanmean: 0.765665 2019-10-09 04:35:27.278: INFO @log_variables: train negative_distance nanmean: 1.408888 2019-10-09 04:35:27.278: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:35:27.278: INFO @log_variables: valid loss mean: 0.455626 2019-10-09 04:35:27.278: INFO @log_variables: valid age_loss mean: 6.661613 2019-10-09 04:35:27.278: INFO @log_variables: valid gender_loss mean: 0.239181 2019-10-09 04:35:27.278: INFO @log_variables: valid matching_loss nanmean: 0.507099 2019-10-09 04:35:27.278: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:35:27.278: INFO @log_variables: valid age_mae mean: 7.144887 2019-10-09 04:35:27.278: INFO @log_variables: valid gender_accuracy mean: 0.922818 2019-10-09 04:35:27.278: INFO @log_variables: valid positive_distance nanmean: 0.784390 2019-10-09 04:35:27.279: INFO @log_variables: valid negative_distance nanmean: 1.370606 2019-10-09 04:35:27.279: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:35:27.279: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:35:29.281: INFO @metrics_hook: valid matching accuracy: 0.8867901900821491, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:35:29.696: INFO @decay_lr : LR updated to `0.00069356145` 2019-10-09 04:35:29.697: INFO @log_profile : T train: 178.149265 2019-10-09 04:35:29.697: INFO @log_profile : T valid: 8.325026 2019-10-09 04:35:29.697: INFO @log_profile : T read data: 1.430826 2019-10-09 04:35:29.698: INFO @log_profile : T hooks: 3.024876 2019-10-09 04:35:29.698: INFO @main_loop : Epoch 73 done 2019-10-09 04:35:29.698: INFO @main_loop : Training epoch 74 2019-10-09 04:38:38.626: INFO @log_variables: train loss mean: 0.298261 2019-10-09 04:38:38.626: INFO @log_variables: train age_loss mean: 4.972062 2019-10-09 04:38:38.626: INFO @log_variables: train gender_loss mean: 0.060391 2019-10-09 04:38:38.626: INFO @log_variables: train matching_loss nanmean: 0.367011 2019-10-09 04:38:38.627: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:38:38.627: INFO @log_variables: train age_mae mean: 5.450920 2019-10-09 04:38:38.627: INFO @log_variables: train gender_accuracy mean: 0.977990 2019-10-09 04:38:38.627: INFO @log_variables: train positive_distance nanmean: 0.766296 2019-10-09 04:38:38.627: INFO @log_variables: train negative_distance nanmean: 1.408828 2019-10-09 04:38:38.627: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:38:38.627: INFO @log_variables: valid loss mean: 0.447746 2019-10-09 04:38:38.627: INFO @log_variables: valid age_loss mean: 6.462074 2019-10-09 04:38:38.627: INFO @log_variables: valid gender_loss mean: 0.233411 2019-10-09 04:38:38.627: INFO @log_variables: valid matching_loss nanmean: 0.508395 2019-10-09 04:38:38.627: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:38:38.627: INFO @log_variables: valid age_mae mean: 6.945863 2019-10-09 04:38:38.627: INFO @log_variables: valid gender_accuracy mean: 0.921162 2019-10-09 04:38:38.627: INFO @log_variables: valid positive_distance nanmean: 0.788006 2019-10-09 04:38:38.627: INFO @log_variables: valid negative_distance nanmean: 1.372134 2019-10-09 04:38:38.627: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:38:38.627: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:38:40.770: INFO @metrics_hook: valid matching accuracy: 0.8884691491275409, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:38:41.181: INFO @decay_lr : LR updated to `0.0006900937` 2019-10-09 04:38:41.182: INFO @log_profile : T train: 178.555360 2019-10-09 04:38:41.182: INFO @log_profile : T valid: 8.313055 2019-10-09 04:38:41.182: INFO @log_profile : T read data: 1.399033 2019-10-09 04:38:41.182: INFO @log_profile : T hooks: 3.131195 2019-10-09 04:38:41.182: INFO @main_loop : Epoch 74 done 2019-10-09 04:38:41.182: INFO @main_loop : Training epoch 75 2019-10-09 04:41:49.174: INFO @log_variables: train loss mean: 0.294823 2019-10-09 04:41:49.175: INFO @log_variables: train age_loss mean: 4.910453 2019-10-09 04:41:49.175: INFO @log_variables: train gender_loss mean: 0.059551 2019-10-09 04:41:49.175: INFO @log_variables: train matching_loss nanmean: 0.363355 2019-10-09 04:41:49.175: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:41:49.175: INFO @log_variables: train age_mae mean: 5.388694 2019-10-09 04:41:49.175: INFO @log_variables: train gender_accuracy mean: 0.978072 2019-10-09 04:41:49.175: INFO @log_variables: train positive_distance nanmean: 0.764926 2019-10-09 04:41:49.175: INFO @log_variables: train negative_distance nanmean: 1.409203 2019-10-09 04:41:49.175: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:41:49.175: INFO @log_variables: valid loss mean: 0.454146 2019-10-09 04:41:49.175: INFO @log_variables: valid age_loss mean: 6.395498 2019-10-09 04:41:49.175: INFO @log_variables: valid gender_loss mean: 0.255908 2019-10-09 04:41:49.175: INFO @log_variables: valid matching_loss nanmean: 0.512396 2019-10-09 04:41:49.175: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:41:49.175: INFO @log_variables: valid age_mae mean: 6.878848 2019-10-09 04:41:49.175: INFO @log_variables: valid gender_accuracy mean: 0.919683 2019-10-09 04:41:49.175: INFO @log_variables: valid positive_distance nanmean: 0.780906 2019-10-09 04:41:49.175: INFO @log_variables: valid negative_distance nanmean: 1.370108 2019-10-09 04:41:49.175: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:41:49.176: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:41:51.769: INFO @metrics_hook: valid matching accuracy: 0.8875097439587456, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:41:52.194: INFO @decay_lr : LR updated to `0.0006866432` 2019-10-09 04:41:52.195: INFO @log_profile : T train: 178.030278 2019-10-09 04:41:52.195: INFO @log_profile : T valid: 8.312952 2019-10-09 04:41:52.195: INFO @log_profile : T read data: 1.017638 2019-10-09 04:41:52.195: INFO @log_profile : T hooks: 3.565842 2019-10-09 04:41:52.195: INFO @main_loop : Epoch 75 done 2019-10-09 04:41:52.195: INFO @main_loop : Training epoch 76 2019-10-09 04:45:00.881: INFO @log_variables: train loss mean: 0.292811 2019-10-09 04:45:00.882: INFO @log_variables: train age_loss mean: 4.893767 2019-10-09 04:45:00.882: INFO @log_variables: train gender_loss mean: 0.056626 2019-10-09 04:45:00.882: INFO @log_variables: train matching_loss nanmean: 0.361711 2019-10-09 04:45:00.882: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:45:00.882: INFO @log_variables: train age_mae mean: 5.371790 2019-10-09 04:45:00.882: INFO @log_variables: train gender_accuracy mean: 0.979502 2019-10-09 04:45:00.882: INFO @log_variables: train positive_distance nanmean: 0.763975 2019-10-09 04:45:00.882: INFO @log_variables: train negative_distance nanmean: 1.409369 2019-10-09 04:45:00.882: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:45:00.882: INFO @log_variables: valid loss mean: 0.457007 2019-10-09 04:45:00.882: INFO @log_variables: valid age_loss mean: 6.564140 2019-10-09 04:45:00.882: INFO @log_variables: valid gender_loss mean: 0.248949 2019-10-09 04:45:00.882: INFO @log_variables: valid matching_loss nanmean: 0.511358 2019-10-09 04:45:00.882: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:45:00.882: INFO @log_variables: valid age_mae mean: 7.047266 2019-10-09 04:45:00.882: INFO @log_variables: valid gender_accuracy mean: 0.922344 2019-10-09 04:45:00.882: INFO @log_variables: valid positive_distance nanmean: 0.783643 2019-10-09 04:45:00.882: INFO @log_variables: valid negative_distance nanmean: 1.369100 2019-10-09 04:45:00.883: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:45:00.883: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:45:03.167: INFO @metrics_hook: valid matching accuracy: 0.8868501529051988, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:45:03.605: INFO @decay_lr : LR updated to `0.00068321` 2019-10-09 04:45:03.607: INFO @log_profile : T train: 178.279034 2019-10-09 04:45:03.607: INFO @log_profile : T valid: 8.303770 2019-10-09 04:45:03.607: INFO @log_profile : T read data: 1.435356 2019-10-09 04:45:03.607: INFO @log_profile : T hooks: 3.307501 2019-10-09 04:45:03.607: INFO @main_loop : Epoch 76 done 2019-10-09 04:45:03.607: INFO @main_loop : Training epoch 77 2019-10-09 04:48:11.728: INFO @log_variables: train loss mean: 0.294690 2019-10-09 04:48:11.728: INFO @log_variables: train age_loss mean: 4.933338 2019-10-09 04:48:11.728: INFO @log_variables: train gender_loss mean: 0.059376 2019-10-09 04:48:11.728: INFO @log_variables: train matching_loss nanmean: 0.360828 2019-10-09 04:48:11.728: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:48:11.728: INFO @log_variables: train age_mae mean: 5.411709 2019-10-09 04:48:11.728: INFO @log_variables: train gender_accuracy mean: 0.977947 2019-10-09 04:48:11.728: INFO @log_variables: train positive_distance nanmean: 0.763302 2019-10-09 04:48:11.728: INFO @log_variables: train negative_distance nanmean: 1.409011 2019-10-09 04:48:11.729: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:48:11.729: INFO @log_variables: valid loss mean: 0.460192 2019-10-09 04:48:11.729: INFO @log_variables: valid age_loss mean: 6.632859 2019-10-09 04:48:11.729: INFO @log_variables: valid gender_loss mean: 0.249320 2019-10-09 04:48:11.729: INFO @log_variables: valid matching_loss nanmean: 0.513990 2019-10-09 04:48:11.729: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:48:11.729: INFO @log_variables: valid age_mae mean: 7.115356 2019-10-09 04:48:11.729: INFO @log_variables: valid gender_accuracy mean: 0.918796 2019-10-09 04:48:11.729: INFO @log_variables: valid positive_distance nanmean: 0.786424 2019-10-09 04:48:11.729: INFO @log_variables: valid negative_distance nanmean: 1.368089 2019-10-09 04:48:11.729: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:48:11.729: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:48:13.934: INFO @metrics_hook: valid matching accuracy: 0.8863104874977514, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:48:14.356: INFO @decay_lr : LR updated to `0.00067979394` 2019-10-09 04:48:14.357: INFO @log_profile : T train: 178.123589 2019-10-09 04:48:14.357: INFO @log_profile : T valid: 8.334423 2019-10-09 04:48:14.357: INFO @log_profile : T read data: 1.015144 2019-10-09 04:48:14.357: INFO @log_profile : T hooks: 3.191094 2019-10-09 04:48:14.357: INFO @main_loop : Epoch 77 done 2019-10-09 04:48:14.357: INFO @main_loop : Training epoch 78 2019-10-09 04:51:23.151: INFO @log_variables: train loss mean: 0.293523 2019-10-09 04:51:23.151: INFO @log_variables: train age_loss mean: 4.930492 2019-10-09 04:51:23.151: INFO @log_variables: train gender_loss mean: 0.056768 2019-10-09 04:51:23.151: INFO @log_variables: train matching_loss nanmean: 0.360105 2019-10-09 04:51:23.151: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:51:23.151: INFO @log_variables: train age_mae mean: 5.408542 2019-10-09 04:51:23.151: INFO @log_variables: train gender_accuracy mean: 0.979339 2019-10-09 04:51:23.151: INFO @log_variables: train positive_distance nanmean: 0.764046 2019-10-09 04:51:23.151: INFO @log_variables: train negative_distance nanmean: 1.409165 2019-10-09 04:51:23.151: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:51:23.151: INFO @log_variables: valid loss mean: 0.449461 2019-10-09 04:51:23.151: INFO @log_variables: valid age_loss mean: 6.453771 2019-10-09 04:51:23.151: INFO @log_variables: valid gender_loss mean: 0.238394 2019-10-09 04:51:23.151: INFO @log_variables: valid matching_loss nanmean: 0.509558 2019-10-09 04:51:23.151: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:51:23.152: INFO @log_variables: valid age_mae mean: 6.936454 2019-10-09 04:51:23.152: INFO @log_variables: valid gender_accuracy mean: 0.920629 2019-10-09 04:51:23.152: INFO @log_variables: valid positive_distance nanmean: 0.779806 2019-10-09 04:51:23.152: INFO @log_variables: valid negative_distance nanmean: 1.366899 2019-10-09 04:51:23.152: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:51:23.152: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:51:25.646: INFO @metrics_hook: valid matching accuracy: 0.8895484799424357, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:51:26.072: INFO @decay_lr : LR updated to `0.00067639496` 2019-10-09 04:51:26.073: INFO @log_profile : T train: 178.291208 2019-10-09 04:51:26.074: INFO @log_profile : T valid: 8.374646 2019-10-09 04:51:26.074: INFO @log_profile : T read data: 1.453547 2019-10-09 04:51:26.074: INFO @log_profile : T hooks: 3.509611 2019-10-09 04:51:26.074: INFO @main_loop : Epoch 78 done 2019-10-09 04:51:26.074: INFO @main_loop : Training epoch 79 2019-10-09 04:54:34.770: INFO @log_variables: train loss mean: 0.291464 2019-10-09 04:54:34.770: INFO @log_variables: train age_loss mean: 4.872158 2019-10-09 04:54:34.770: INFO @log_variables: train gender_loss mean: 0.057503 2019-10-09 04:54:34.770: INFO @log_variables: train matching_loss nanmean: 0.358821 2019-10-09 04:54:34.770: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:54:34.770: INFO @log_variables: train age_mae mean: 5.350204 2019-10-09 04:54:34.771: INFO @log_variables: train gender_accuracy mean: 0.978834 2019-10-09 04:54:34.771: INFO @log_variables: train positive_distance nanmean: 0.762196 2019-10-09 04:54:34.771: INFO @log_variables: train negative_distance nanmean: 1.409150 2019-10-09 04:54:34.771: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:54:34.771: INFO @log_variables: valid loss mean: 0.449503 2019-10-09 04:54:34.771: INFO @log_variables: valid age_loss mean: 6.493964 2019-10-09 04:54:34.771: INFO @log_variables: valid gender_loss mean: 0.235803 2019-10-09 04:54:34.771: INFO @log_variables: valid matching_loss nanmean: 0.508259 2019-10-09 04:54:34.771: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:54:34.771: INFO @log_variables: valid age_mae mean: 6.976874 2019-10-09 04:54:34.771: INFO @log_variables: valid gender_accuracy mean: 0.919032 2019-10-09 04:54:34.771: INFO @log_variables: valid positive_distance nanmean: 0.782820 2019-10-09 04:54:34.771: INFO @log_variables: valid negative_distance nanmean: 1.370673 2019-10-09 04:54:34.771: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:54:34.771: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:54:37.097: INFO @metrics_hook: valid matching accuracy: 0.8870300413743479, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:54:37.510: INFO @decay_lr : LR updated to `0.000673013` 2019-10-09 04:54:37.511: INFO @log_profile : T train: 178.247644 2019-10-09 04:54:37.511: INFO @log_profile : T valid: 8.308690 2019-10-09 04:54:37.511: INFO @log_profile : T read data: 1.486971 2019-10-09 04:54:37.511: INFO @log_profile : T hooks: 3.309181 2019-10-09 04:54:37.511: INFO @main_loop : Epoch 79 done 2019-10-09 04:54:37.511: INFO @main_loop : Training epoch 80 2019-10-09 04:57:45.698: INFO @log_variables: train loss mean: 0.291774 2019-10-09 04:57:45.698: INFO @log_variables: train age_loss mean: 4.885333 2019-10-09 04:57:45.698: INFO @log_variables: train gender_loss mean: 0.057506 2019-10-09 04:57:45.699: INFO @log_variables: train matching_loss nanmean: 0.358459 2019-10-09 04:57:45.699: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 04:57:45.699: INFO @log_variables: train age_mae mean: 5.363274 2019-10-09 04:57:45.699: INFO @log_variables: train gender_accuracy mean: 0.979133 2019-10-09 04:57:45.699: INFO @log_variables: train positive_distance nanmean: 0.762948 2019-10-09 04:57:45.699: INFO @log_variables: train negative_distance nanmean: 1.409392 2019-10-09 04:57:45.699: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 04:57:45.699: INFO @log_variables: valid loss mean: 0.456556 2019-10-09 04:57:45.699: INFO @log_variables: valid age_loss mean: 6.664652 2019-10-09 04:57:45.699: INFO @log_variables: valid gender_loss mean: 0.240717 2019-10-09 04:57:45.699: INFO @log_variables: valid matching_loss nanmean: 0.508143 2019-10-09 04:57:45.699: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 04:57:45.699: INFO @log_variables: valid age_mae mean: 7.149332 2019-10-09 04:57:45.699: INFO @log_variables: valid gender_accuracy mean: 0.921576 2019-10-09 04:57:45.699: INFO @log_variables: valid positive_distance nanmean: 0.780759 2019-10-09 04:57:45.699: INFO @log_variables: valid negative_distance nanmean: 1.371201 2019-10-09 04:57:45.699: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 04:57:45.699: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 04:57:47.970: INFO @metrics_hook: valid matching accuracy: 0.8897883312346345, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 04:57:48.389: INFO @decay_lr : LR updated to `0.00066964794` 2019-10-09 04:57:48.390: INFO @log_profile : T train: 178.209905 2019-10-09 04:57:48.390: INFO @log_profile : T valid: 8.312711 2019-10-09 04:57:48.390: INFO @log_profile : T read data: 1.023484 2019-10-09 04:57:48.390: INFO @log_profile : T hooks: 3.248142 2019-10-09 04:57:48.390: INFO @main_loop : Epoch 80 done 2019-10-09 04:57:48.390: INFO @main_loop : Training epoch 81 2019-10-09 05:00:56.940: INFO @log_variables: train loss mean: 0.289669 2019-10-09 05:00:56.940: INFO @log_variables: train age_loss mean: 4.834036 2019-10-09 05:00:56.940: INFO @log_variables: train gender_loss mean: 0.056125 2019-10-09 05:00:56.940: INFO @log_variables: train matching_loss nanmean: 0.358446 2019-10-09 05:00:56.940: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:00:56.940: INFO @log_variables: train age_mae mean: 5.312082 2019-10-09 05:00:56.940: INFO @log_variables: train gender_accuracy mean: 0.979277 2019-10-09 05:00:56.940: INFO @log_variables: train positive_distance nanmean: 0.761799 2019-10-09 05:00:56.940: INFO @log_variables: train negative_distance nanmean: 1.409087 2019-10-09 05:00:56.940: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:00:56.940: INFO @log_variables: valid loss mean: 0.450073 2019-10-09 05:00:56.940: INFO @log_variables: valid age_loss mean: 6.475231 2019-10-09 05:00:56.940: INFO @log_variables: valid gender_loss mean: 0.245317 2019-10-09 05:00:56.940: INFO @log_variables: valid matching_loss nanmean: 0.502385 2019-10-09 05:00:56.940: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:00:56.940: INFO @log_variables: valid age_mae mean: 6.960090 2019-10-09 05:00:56.941: INFO @log_variables: valid gender_accuracy mean: 0.922167 2019-10-09 05:00:56.941: INFO @log_variables: valid positive_distance nanmean: 0.786044 2019-10-09 05:00:56.941: INFO @log_variables: valid negative_distance nanmean: 1.373502 2019-10-09 05:00:56.941: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:00:56.941: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:00:58.971: INFO @metrics_hook: valid matching accuracy: 0.8884091863044912, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:00:59.382: INFO @decay_lr : LR updated to `0.0006662997` 2019-10-09 05:00:59.383: INFO @log_profile : T train: 178.060186 2019-10-09 05:00:59.384: INFO @log_profile : T valid: 8.458911 2019-10-09 05:00:59.384: INFO @log_profile : T read data: 1.402714 2019-10-09 05:00:59.384: INFO @log_profile : T hooks: 2.989171 2019-10-09 05:00:59.384: INFO @main_loop : Epoch 81 done 2019-10-09 05:00:59.384: INFO @main_loop : Training epoch 82 2019-10-09 05:04:07.934: INFO @log_variables: train loss mean: 0.290116 2019-10-09 05:04:07.934: INFO @log_variables: train age_loss mean: 4.828892 2019-10-09 05:04:07.934: INFO @log_variables: train gender_loss mean: 0.058341 2019-10-09 05:04:07.934: INFO @log_variables: train matching_loss nanmean: 0.358131 2019-10-09 05:04:07.934: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:04:07.934: INFO @log_variables: train age_mae mean: 5.306935 2019-10-09 05:04:07.934: INFO @log_variables: train gender_accuracy mean: 0.979066 2019-10-09 05:04:07.934: INFO @log_variables: train positive_distance nanmean: 0.762354 2019-10-09 05:04:07.934: INFO @log_variables: train negative_distance nanmean: 1.408887 2019-10-09 05:04:07.934: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:04:07.934: INFO @log_variables: valid loss mean: 0.438717 2019-10-09 05:04:07.934: INFO @log_variables: valid age_loss mean: 6.268750 2019-10-09 05:04:07.934: INFO @log_variables: valid gender_loss mean: 0.226018 2019-10-09 05:04:07.934: INFO @log_variables: valid matching_loss nanmean: 0.507129 2019-10-09 05:04:07.934: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:04:07.935: INFO @log_variables: valid age_mae mean: 6.750546 2019-10-09 05:04:07.935: INFO @log_variables: valid gender_accuracy mean: 0.923705 2019-10-09 05:04:07.935: INFO @log_variables: valid positive_distance nanmean: 0.781904 2019-10-09 05:04:07.935: INFO @log_variables: valid negative_distance nanmean: 1.371229 2019-10-09 05:04:07.935: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:04:07.935: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:04:10.623: INFO @metrics_hook: valid matching accuracy: 0.8860706362055526, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:04:11.052: INFO @decay_lr : LR updated to `0.0006629682` 2019-10-09 05:04:11.054: INFO @log_profile : T train: 178.120378 2019-10-09 05:04:11.054: INFO @log_profile : T valid: 8.322386 2019-10-09 05:04:11.054: INFO @log_profile : T read data: 1.434551 2019-10-09 05:04:11.054: INFO @log_profile : T hooks: 3.706252 2019-10-09 05:04:11.054: INFO @main_loop : Epoch 82 done 2019-10-09 05:04:11.054: INFO @main_loop : Training epoch 83 2019-10-09 05:07:19.212: INFO @log_variables: train loss mean: 0.288289 2019-10-09 05:07:19.212: INFO @log_variables: train age_loss mean: 4.822403 2019-10-09 05:07:19.213: INFO @log_variables: train gender_loss mean: 0.056366 2019-10-09 05:07:19.213: INFO @log_variables: train matching_loss nanmean: 0.355088 2019-10-09 05:07:19.213: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:07:19.213: INFO @log_variables: train age_mae mean: 5.300212 2019-10-09 05:07:19.213: INFO @log_variables: train gender_accuracy mean: 0.979919 2019-10-09 05:07:19.213: INFO @log_variables: train positive_distance nanmean: 0.761487 2019-10-09 05:07:19.213: INFO @log_variables: train negative_distance nanmean: 1.409031 2019-10-09 05:07:19.213: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:07:19.213: INFO @log_variables: valid loss mean: 0.449547 2019-10-09 05:07:19.213: INFO @log_variables: valid age_loss mean: 6.282683 2019-10-09 05:07:19.213: INFO @log_variables: valid gender_loss mean: 0.262583 2019-10-09 05:07:19.213: INFO @log_variables: valid matching_loss nanmean: 0.502744 2019-10-09 05:07:19.213: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:07:19.213: INFO @log_variables: valid age_mae mean: 6.766371 2019-10-09 05:07:19.213: INFO @log_variables: valid gender_accuracy mean: 0.916489 2019-10-09 05:07:19.214: INFO @log_variables: valid positive_distance nanmean: 0.778141 2019-10-09 05:07:19.214: INFO @log_variables: valid negative_distance nanmean: 1.368785 2019-10-09 05:07:19.214: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:07:19.214: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:07:21.439: INFO @metrics_hook: valid matching accuracy: 0.8887090004197398, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:07:21.849: INFO @decay_lr : LR updated to `0.00065965333` 2019-10-09 05:07:21.851: INFO @log_profile : T train: 178.105242 2019-10-09 05:07:21.851: INFO @log_profile : T valid: 8.338924 2019-10-09 05:07:21.851: INFO @log_profile : T read data: 1.045874 2019-10-09 05:07:21.851: INFO @log_profile : T hooks: 3.219196 2019-10-09 05:07:21.851: INFO @main_loop : Epoch 83 done 2019-10-09 05:07:21.851: INFO @main_loop : Training epoch 84 2019-10-09 05:10:30.396: INFO @log_variables: train loss mean: 0.288793 2019-10-09 05:10:30.396: INFO @log_variables: train age_loss mean: 4.798523 2019-10-09 05:10:30.396: INFO @log_variables: train gender_loss mean: 0.056592 2019-10-09 05:10:30.397: INFO @log_variables: train matching_loss nanmean: 0.358814 2019-10-09 05:10:30.397: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:10:30.397: INFO @log_variables: train age_mae mean: 5.276344 2019-10-09 05:10:30.397: INFO @log_variables: train gender_accuracy mean: 0.978867 2019-10-09 05:10:30.397: INFO @log_variables: train positive_distance nanmean: 0.764066 2019-10-09 05:10:30.397: INFO @log_variables: train negative_distance nanmean: 1.409126 2019-10-09 05:10:30.397: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:10:30.397: INFO @log_variables: valid loss mean: 0.446860 2019-10-09 05:10:30.397: INFO @log_variables: valid age_loss mean: 6.403913 2019-10-09 05:10:30.397: INFO @log_variables: valid gender_loss mean: 0.234056 2019-10-09 05:10:30.397: INFO @log_variables: valid matching_loss nanmean: 0.510820 2019-10-09 05:10:30.397: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:10:30.397: INFO @log_variables: valid age_mae mean: 6.887042 2019-10-09 05:10:30.397: INFO @log_variables: valid gender_accuracy mean: 0.922995 2019-10-09 05:10:30.397: INFO @log_variables: valid positive_distance nanmean: 0.785138 2019-10-09 05:10:30.397: INFO @log_variables: valid negative_distance nanmean: 1.371869 2019-10-09 05:10:30.397: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:10:30.397: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:10:32.497: INFO @metrics_hook: valid matching accuracy: 0.8887689632427894, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:10:32.922: INFO @decay_lr : LR updated to `0.00065635506` 2019-10-09 05:10:32.923: INFO @log_profile : T train: 178.126951 2019-10-09 05:10:32.923: INFO @log_profile : T valid: 8.331532 2019-10-09 05:10:32.923: INFO @log_profile : T read data: 1.427782 2019-10-09 05:10:32.923: INFO @log_profile : T hooks: 3.100889 2019-10-09 05:10:32.923: INFO @main_loop : Epoch 84 done 2019-10-09 05:10:32.924: INFO @main_loop : Training epoch 85 2019-10-09 05:13:41.314: INFO @log_variables: train loss mean: 0.287930 2019-10-09 05:13:41.314: INFO @log_variables: train age_loss mean: 4.809519 2019-10-09 05:13:41.314: INFO @log_variables: train gender_loss mean: 0.055431 2019-10-09 05:13:41.314: INFO @log_variables: train matching_loss nanmean: 0.356201 2019-10-09 05:13:41.314: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:13:41.314: INFO @log_variables: train age_mae mean: 5.287600 2019-10-09 05:13:41.314: INFO @log_variables: train gender_accuracy mean: 0.980099 2019-10-09 05:13:41.314: INFO @log_variables: train positive_distance nanmean: 0.761203 2019-10-09 05:13:41.314: INFO @log_variables: train negative_distance nanmean: 1.408993 2019-10-09 05:13:41.314: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:13:41.314: INFO @log_variables: valid loss mean: 0.440895 2019-10-09 05:13:41.315: INFO @log_variables: valid age_loss mean: 6.350727 2019-10-09 05:13:41.315: INFO @log_variables: valid gender_loss mean: 0.225005 2019-10-09 05:13:41.315: INFO @log_variables: valid matching_loss nanmean: 0.506698 2019-10-09 05:13:41.315: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:13:41.315: INFO @log_variables: valid age_mae mean: 6.833775 2019-10-09 05:13:41.315: INFO @log_variables: valid gender_accuracy mean: 0.923882 2019-10-09 05:13:41.315: INFO @log_variables: valid positive_distance nanmean: 0.783749 2019-10-09 05:13:41.315: INFO @log_variables: valid negative_distance nanmean: 1.371712 2019-10-09 05:13:41.315: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:13:41.315: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:13:43.436: INFO @metrics_hook: valid matching accuracy: 0.8867901900821491, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:13:43.860: INFO @decay_lr : LR updated to `0.0006530733` 2019-10-09 05:13:43.861: INFO @log_profile : T train: 178.404970 2019-10-09 05:13:43.861: INFO @log_profile : T valid: 8.306094 2019-10-09 05:13:43.861: INFO @log_profile : T read data: 1.021814 2019-10-09 05:13:43.861: INFO @log_profile : T hooks: 3.119624 2019-10-09 05:13:43.861: INFO @main_loop : Epoch 85 done 2019-10-09 05:13:43.861: INFO @main_loop : Training epoch 86 2019-10-09 05:16:52.692: INFO @log_variables: train loss mean: 0.286382 2019-10-09 05:16:52.692: INFO @log_variables: train age_loss mean: 4.759035 2019-10-09 05:16:52.692: INFO @log_variables: train gender_loss mean: 0.055758 2019-10-09 05:16:52.692: INFO @log_variables: train matching_loss nanmean: 0.356123 2019-10-09 05:16:52.692: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:16:52.692: INFO @log_variables: train age_mae mean: 5.236308 2019-10-09 05:16:52.692: INFO @log_variables: train gender_accuracy mean: 0.979932 2019-10-09 05:16:52.692: INFO @log_variables: train positive_distance nanmean: 0.761577 2019-10-09 05:16:52.692: INFO @log_variables: train negative_distance nanmean: 1.409010 2019-10-09 05:16:52.692: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:16:52.692: INFO @log_variables: valid loss mean: 0.448891 2019-10-09 05:16:52.692: INFO @log_variables: valid age_loss mean: 6.381417 2019-10-09 05:16:52.692: INFO @log_variables: valid gender_loss mean: 0.243516 2019-10-09 05:16:52.693: INFO @log_variables: valid matching_loss nanmean: 0.509904 2019-10-09 05:16:52.693: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:16:52.693: INFO @log_variables: valid age_mae mean: 6.862527 2019-10-09 05:16:52.693: INFO @log_variables: valid gender_accuracy mean: 0.923882 2019-10-09 05:16:52.693: INFO @log_variables: valid positive_distance nanmean: 0.778858 2019-10-09 05:16:52.693: INFO @log_variables: valid negative_distance nanmean: 1.368749 2019-10-09 05:16:52.693: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:16:52.693: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:16:54.886: INFO @metrics_hook: valid matching accuracy: 0.8883492234814415, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:16:55.335: INFO @decay_lr : LR updated to `0.0006498079` 2019-10-09 05:16:55.336: INFO @log_profile : T train: 178.379844 2019-10-09 05:16:55.336: INFO @log_profile : T valid: 8.361298 2019-10-09 05:16:55.336: INFO @log_profile : T read data: 1.421251 2019-10-09 05:16:55.336: INFO @log_profile : T hooks: 3.226300 2019-10-09 05:16:55.336: INFO @main_loop : Epoch 86 done 2019-10-09 05:16:55.336: INFO @main_loop : Training epoch 87 2019-10-09 05:20:04.255: INFO @log_variables: train loss mean: 0.285266 2019-10-09 05:20:04.255: INFO @log_variables: train age_loss mean: 4.752554 2019-10-09 05:20:04.255: INFO @log_variables: train gender_loss mean: 0.055243 2019-10-09 05:20:04.255: INFO @log_variables: train matching_loss nanmean: 0.353826 2019-10-09 05:20:04.255: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:20:04.255: INFO @log_variables: train age_mae mean: 5.230142 2019-10-09 05:20:04.255: INFO @log_variables: train gender_accuracy mean: 0.979668 2019-10-09 05:20:04.255: INFO @log_variables: train positive_distance nanmean: 0.758847 2019-10-09 05:20:04.255: INFO @log_variables: train negative_distance nanmean: 1.408676 2019-10-09 05:20:04.255: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:20:04.255: INFO @log_variables: valid loss mean: 0.448940 2019-10-09 05:20:04.255: INFO @log_variables: valid age_loss mean: 6.250320 2019-10-09 05:20:04.255: INFO @log_variables: valid gender_loss mean: 0.256288 2019-10-09 05:20:04.255: INFO @log_variables: valid matching_loss nanmean: 0.510392 2019-10-09 05:20:04.256: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:20:04.256: INFO @log_variables: valid age_mae mean: 6.731345 2019-10-09 05:20:04.256: INFO @log_variables: valid gender_accuracy mean: 0.924001 2019-10-09 05:20:04.256: INFO @log_variables: valid positive_distance nanmean: 0.780541 2019-10-09 05:20:04.256: INFO @log_variables: valid negative_distance nanmean: 1.372411 2019-10-09 05:20:04.256: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:20:04.256: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:20:06.537: INFO @metrics_hook: valid matching accuracy: 0.8879294837200935, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:20:06.960: INFO @decay_lr : LR updated to `0.0006465589` 2019-10-09 05:20:06.961: INFO @log_profile : T train: 178.432620 2019-10-09 05:20:06.961: INFO @log_profile : T valid: 8.380720 2019-10-09 05:20:06.961: INFO @log_profile : T read data: 1.441499 2019-10-09 05:20:06.962: INFO @log_profile : T hooks: 3.283783 2019-10-09 05:20:06.962: INFO @main_loop : Epoch 87 done 2019-10-09 05:20:06.962: INFO @main_loop : Training epoch 88 2019-10-09 05:23:15.454: INFO @log_variables: train loss mean: 0.286140 2019-10-09 05:23:15.454: INFO @log_variables: train age_loss mean: 4.778987 2019-10-09 05:23:15.454: INFO @log_variables: train gender_loss mean: 0.053380 2019-10-09 05:23:15.454: INFO @log_variables: train matching_loss nanmean: 0.355754 2019-10-09 05:23:15.454: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:23:15.454: INFO @log_variables: train age_mae mean: 5.256617 2019-10-09 05:23:15.455: INFO @log_variables: train gender_accuracy mean: 0.981013 2019-10-09 05:23:15.455: INFO @log_variables: train positive_distance nanmean: 0.761240 2019-10-09 05:23:15.455: INFO @log_variables: train negative_distance nanmean: 1.409011 2019-10-09 05:23:15.455: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:23:15.455: INFO @log_variables: valid loss mean: 0.458824 2019-10-09 05:23:15.455: INFO @log_variables: valid age_loss mean: 6.497894 2019-10-09 05:23:15.455: INFO @log_variables: valid gender_loss mean: 0.265389 2019-10-09 05:23:15.455: INFO @log_variables: valid matching_loss nanmean: 0.507175 2019-10-09 05:23:15.455: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:23:15.455: INFO @log_variables: valid age_mae mean: 6.980854 2019-10-09 05:23:15.455: INFO @log_variables: valid gender_accuracy mean: 0.919387 2019-10-09 05:23:15.455: INFO @log_variables: valid positive_distance nanmean: 0.784734 2019-10-09 05:23:15.455: INFO @log_variables: valid negative_distance nanmean: 1.373588 2019-10-09 05:23:15.455: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:23:15.455: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:23:17.848: INFO @metrics_hook: valid matching accuracy: 0.8887689632427894, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:23:18.272: INFO @decay_lr : LR updated to `0.0006433261` 2019-10-09 05:23:18.273: INFO @log_profile : T train: 178.445833 2019-10-09 05:23:18.273: INFO @log_profile : T valid: 8.340322 2019-10-09 05:23:18.273: INFO @log_profile : T read data: 1.041029 2019-10-09 05:23:18.273: INFO @log_profile : T hooks: 3.397655 2019-10-09 05:23:18.273: INFO @main_loop : Epoch 88 done 2019-10-09 05:23:18.273: INFO @main_loop : Training epoch 89 2019-10-09 05:26:26.824: INFO @log_variables: train loss mean: 0.285912 2019-10-09 05:26:26.824: INFO @log_variables: train age_loss mean: 4.742632 2019-10-09 05:26:26.824: INFO @log_variables: train gender_loss mean: 0.059899 2019-10-09 05:26:26.824: INFO @log_variables: train matching_loss nanmean: 0.352164 2019-10-09 05:26:26.824: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:26:26.824: INFO @log_variables: train age_mae mean: 5.220153 2019-10-09 05:26:26.824: INFO @log_variables: train gender_accuracy mean: 0.978444 2019-10-09 05:26:26.824: INFO @log_variables: train positive_distance nanmean: 0.760351 2019-10-09 05:26:26.824: INFO @log_variables: train negative_distance nanmean: 1.408885 2019-10-09 05:26:26.824: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:26:26.824: INFO @log_variables: valid loss mean: 0.446081 2019-10-09 05:26:26.825: INFO @log_variables: valid age_loss mean: 6.491001 2019-10-09 05:26:26.825: INFO @log_variables: valid gender_loss mean: 0.222559 2019-10-09 05:26:26.825: INFO @log_variables: valid matching_loss nanmean: 0.511191 2019-10-09 05:26:26.825: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:26:26.825: INFO @log_variables: valid age_mae mean: 6.974882 2019-10-09 05:26:26.825: INFO @log_variables: valid gender_accuracy mean: 0.925124 2019-10-09 05:26:26.825: INFO @log_variables: valid positive_distance nanmean: 0.782292 2019-10-09 05:26:26.825: INFO @log_variables: valid negative_distance nanmean: 1.370847 2019-10-09 05:26:26.825: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:26:26.825: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:26:28.991: INFO @metrics_hook: valid matching accuracy: 0.8852311566828567, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:26:29.410: INFO @decay_lr : LR updated to `0.0006401095` 2019-10-09 05:26:29.412: INFO @log_profile : T train: 178.209273 2019-10-09 05:26:29.412: INFO @log_profile : T valid: 8.313340 2019-10-09 05:26:29.412: INFO @log_profile : T read data: 1.385835 2019-10-09 05:26:29.412: INFO @log_profile : T hooks: 3.144924 2019-10-09 05:26:29.412: INFO @main_loop : Epoch 89 done 2019-10-09 05:26:29.412: INFO @main_loop : Training epoch 90 2019-10-09 05:29:38.187: INFO @log_variables: train loss mean: 0.283604 2019-10-09 05:29:38.187: INFO @log_variables: train age_loss mean: 4.740210 2019-10-09 05:29:38.187: INFO @log_variables: train gender_loss mean: 0.051666 2019-10-09 05:29:38.187: INFO @log_variables: train matching_loss nanmean: 0.353486 2019-10-09 05:29:38.187: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:29:38.187: INFO @log_variables: train age_mae mean: 5.217782 2019-10-09 05:29:38.187: INFO @log_variables: train gender_accuracy mean: 0.981401 2019-10-09 05:29:38.187: INFO @log_variables: train positive_distance nanmean: 0.759941 2019-10-09 05:29:38.187: INFO @log_variables: train negative_distance nanmean: 1.408571 2019-10-09 05:29:38.187: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:29:38.187: INFO @log_variables: valid loss mean: 0.451061 2019-10-09 05:29:38.187: INFO @log_variables: valid age_loss mean: 6.405263 2019-10-09 05:29:38.188: INFO @log_variables: valid gender_loss mean: 0.254100 2019-10-09 05:29:38.188: INFO @log_variables: valid matching_loss nanmean: 0.503664 2019-10-09 05:29:38.188: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:29:38.188: INFO @log_variables: valid age_mae mean: 6.888998 2019-10-09 05:29:38.188: INFO @log_variables: valid gender_accuracy mean: 0.921635 2019-10-09 05:29:38.188: INFO @log_variables: valid positive_distance nanmean: 0.781930 2019-10-09 05:29:38.188: INFO @log_variables: valid negative_distance nanmean: 1.369505 2019-10-09 05:29:38.188: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:29:38.188: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:29:40.697: INFO @metrics_hook: valid matching accuracy: 0.8897283684115849, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:29:41.144: INFO @decay_lr : LR updated to `0.0006369089` 2019-10-09 05:29:41.146: INFO @log_profile : T train: 178.309609 2019-10-09 05:29:41.146: INFO @log_profile : T valid: 8.341194 2019-10-09 05:29:41.146: INFO @log_profile : T read data: 1.459675 2019-10-09 05:29:41.146: INFO @log_profile : T hooks: 3.537425 2019-10-09 05:29:41.146: INFO @main_loop : Epoch 90 done 2019-10-09 05:29:41.146: INFO @main_loop : Training epoch 91 2019-10-09 05:32:49.373: INFO @log_variables: train loss mean: 0.283049 2019-10-09 05:32:49.373: INFO @log_variables: train age_loss mean: 4.746235 2019-10-09 05:32:49.373: INFO @log_variables: train gender_loss mean: 0.052637 2019-10-09 05:32:49.374: INFO @log_variables: train matching_loss nanmean: 0.350190 2019-10-09 05:32:49.374: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:32:49.374: INFO @log_variables: train age_mae mean: 5.223724 2019-10-09 05:32:49.374: INFO @log_variables: train gender_accuracy mean: 0.981168 2019-10-09 05:32:49.374: INFO @log_variables: train positive_distance nanmean: 0.758350 2019-10-09 05:32:49.374: INFO @log_variables: train negative_distance nanmean: 1.408568 2019-10-09 05:32:49.374: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:32:49.374: INFO @log_variables: valid loss mean: 0.442854 2019-10-09 05:32:49.374: INFO @log_variables: valid age_loss mean: 6.257829 2019-10-09 05:32:49.374: INFO @log_variables: valid gender_loss mean: 0.245834 2019-10-09 05:32:49.374: INFO @log_variables: valid matching_loss nanmean: 0.501231 2019-10-09 05:32:49.374: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:32:49.374: INFO @log_variables: valid age_mae mean: 6.740678 2019-10-09 05:32:49.374: INFO @log_variables: valid gender_accuracy mean: 0.923882 2019-10-09 05:32:49.374: INFO @log_variables: valid positive_distance nanmean: 0.781259 2019-10-09 05:32:49.374: INFO @log_variables: valid negative_distance nanmean: 1.367495 2019-10-09 05:32:49.374: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:32:49.374: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:32:51.452: INFO @metrics_hook: valid matching accuracy: 0.8891287401810877, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:32:51.863: INFO @decay_lr : LR updated to `0.0006337244` 2019-10-09 05:32:51.864: INFO @log_profile : T train: 178.223844 2019-10-09 05:32:51.864: INFO @log_profile : T valid: 8.307953 2019-10-09 05:32:51.864: INFO @log_profile : T read data: 1.027205 2019-10-09 05:32:51.864: INFO @log_profile : T hooks: 3.072121 2019-10-09 05:32:51.864: INFO @main_loop : Epoch 91 done 2019-10-09 05:32:51.864: INFO @main_loop : Training epoch 92 2019-10-09 05:36:00.507: INFO @log_variables: train loss mean: 0.282369 2019-10-09 05:36:00.508: INFO @log_variables: train age_loss mean: 4.700933 2019-10-09 05:36:00.508: INFO @log_variables: train gender_loss mean: 0.054145 2019-10-09 05:36:00.508: INFO @log_variables: train matching_loss nanmean: 0.351107 2019-10-09 05:36:00.508: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:36:00.508: INFO @log_variables: train age_mae mean: 5.178280 2019-10-09 05:36:00.508: INFO @log_variables: train gender_accuracy mean: 0.980127 2019-10-09 05:36:00.508: INFO @log_variables: train positive_distance nanmean: 0.759136 2019-10-09 05:36:00.508: INFO @log_variables: train negative_distance nanmean: 1.408746 2019-10-09 05:36:00.508: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:36:00.508: INFO @log_variables: valid loss mean: 0.456032 2019-10-09 05:36:00.508: INFO @log_variables: valid age_loss mean: 6.448592 2019-10-09 05:36:00.508: INFO @log_variables: valid gender_loss mean: 0.257880 2019-10-09 05:36:00.508: INFO @log_variables: valid matching_loss nanmean: 0.510959 2019-10-09 05:36:00.508: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:36:00.508: INFO @log_variables: valid age_mae mean: 6.931358 2019-10-09 05:36:00.508: INFO @log_variables: valid gender_accuracy mean: 0.916075 2019-10-09 05:36:00.508: INFO @log_variables: valid positive_distance nanmean: 0.780118 2019-10-09 05:36:00.508: INFO @log_variables: valid negative_distance nanmean: 1.368438 2019-10-09 05:36:00.509: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:36:00.509: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:36:02.548: INFO @metrics_hook: valid matching accuracy: 0.8863104874977514, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:36:02.979: INFO @decay_lr : LR updated to `0.00063055573` 2019-10-09 05:36:02.980: INFO @log_profile : T train: 178.217030 2019-10-09 05:36:02.980: INFO @log_profile : T valid: 8.298298 2019-10-09 05:36:02.980: INFO @log_profile : T read data: 1.466835 2019-10-09 05:36:02.980: INFO @log_profile : T hooks: 3.047507 2019-10-09 05:36:02.980: INFO @main_loop : Epoch 92 done 2019-10-09 05:36:02.980: INFO @main_loop : Training epoch 93 2019-10-09 05:39:11.255: INFO @log_variables: train loss mean: 0.280963 2019-10-09 05:39:11.255: INFO @log_variables: train age_loss mean: 4.690896 2019-10-09 05:39:11.255: INFO @log_variables: train gender_loss mean: 0.050759 2019-10-09 05:39:11.255: INFO @log_variables: train matching_loss nanmean: 0.351135 2019-10-09 05:39:11.255: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 05:39:11.255: INFO @log_variables: train age_mae mean: 5.168188 2019-10-09 05:39:11.255: INFO @log_variables: train gender_accuracy mean: 0.981142 2019-10-09 05:39:11.255: INFO @log_variables: train positive_distance nanmean: 0.760016 2019-10-09 05:39:11.256: INFO @log_variables: train negative_distance nanmean: 1.409147 2019-10-09 05:39:11.256: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:39:11.256: INFO @log_variables: valid loss mean: 0.443525 2019-10-09 05:39:11.256: INFO @log_variables: valid age_loss mean: 6.256837 2019-10-09 05:39:11.256: INFO @log_variables: valid gender_loss mean: 0.241445 2019-10-09 05:39:11.256: INFO @log_variables: valid matching_loss nanmean: 0.507798 2019-10-09 05:39:11.256: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:39:11.256: INFO @log_variables: valid age_mae mean: 6.737661 2019-10-09 05:39:11.256: INFO @log_variables: valid gender_accuracy mean: 0.925538 2019-10-09 05:39:11.256: INFO @log_variables: valid positive_distance nanmean: 0.779631 2019-10-09 05:39:11.256: INFO @log_variables: valid negative_distance nanmean: 1.369764 2019-10-09 05:39:11.256: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:39:11.256: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:39:13.492: INFO @metrics_hook: valid matching accuracy: 0.8917071415722252, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:39:13.907: INFO @decay_lr : LR updated to `0.000627403` 2019-10-09 05:39:14.600: INFO @model : Quantizing and saving the model 2019-10-09 05:39:15.419: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 05:39:15.426: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 05:39:15.431: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 05:39:15.436: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 05:39:15.442: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 05:39:15.447: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 05:39:15.452: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 05:39:15.458: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 05:39:15.463: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 05:39:15.468: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 05:39:15.473: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 05:39:15.478: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 05:39:15.483: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 05:39:15.488: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 05:39:15.494: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 05:39:15.499: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 05:39:15.505: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 05:39:15.510: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 05:39:15.515: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 05:39:15.520: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 05:39:15.525: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 05:39:15.530: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 05:39:15.536: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 05:39:15.541: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 05:39:15.546: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 05:39:15.551: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 05:39:15.556: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 05:39:15.563: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 05:39:27.960: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 05:39:28.231: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 05:39:28.249: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 05:39:30.208: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 05:39:30.251: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 05:39:30.255: INFO @log_profile : T train: 178.067358 2019-10-09 05:39:30.255: INFO @log_profile : T valid: 8.490331 2019-10-09 05:39:30.256: INFO @log_profile : T read data: 1.040898 2019-10-09 05:39:30.256: INFO @log_profile : T hooks: 19.589389 2019-10-09 05:39:30.257: INFO @main_loop : Epoch 93 done 2019-10-09 05:39:30.257: INFO @main_loop : Training epoch 94 2019-10-09 05:42:38.819: INFO @log_variables: train loss mean: 0.279285 2019-10-09 05:42:38.819: INFO @log_variables: train age_loss mean: 4.684759 2019-10-09 05:42:38.819: INFO @log_variables: train gender_loss mean: 0.051437 2019-10-09 05:42:38.819: INFO @log_variables: train matching_loss nanmean: 0.345870 2019-10-09 05:42:38.819: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:42:38.819: INFO @log_variables: train age_mae mean: 5.162334 2019-10-09 05:42:38.819: INFO @log_variables: train gender_accuracy mean: 0.981400 2019-10-09 05:42:38.819: INFO @log_variables: train positive_distance nanmean: 0.757028 2019-10-09 05:42:38.819: INFO @log_variables: train negative_distance nanmean: 1.409262 2019-10-09 05:42:38.819: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:42:38.819: INFO @log_variables: valid loss mean: 0.453236 2019-10-09 05:42:38.819: INFO @log_variables: valid age_loss mean: 6.424221 2019-10-09 05:42:38.819: INFO @log_variables: valid gender_loss mean: 0.259067 2019-10-09 05:42:38.820: INFO @log_variables: valid matching_loss nanmean: 0.503542 2019-10-09 05:42:38.820: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:42:38.820: INFO @log_variables: valid age_mae mean: 6.907037 2019-10-09 05:42:38.820: INFO @log_variables: valid gender_accuracy mean: 0.922049 2019-10-09 05:42:38.820: INFO @log_variables: valid positive_distance nanmean: 0.786740 2019-10-09 05:42:38.820: INFO @log_variables: valid negative_distance nanmean: 1.372228 2019-10-09 05:42:38.820: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:42:38.820: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:42:41.267: INFO @metrics_hook: valid matching accuracy: 0.8900881453498831, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:42:41.701: INFO @decay_lr : LR updated to `0.00062426593` 2019-10-09 05:42:41.702: INFO @log_profile : T train: 178.125876 2019-10-09 05:42:41.702: INFO @log_profile : T valid: 8.328251 2019-10-09 05:42:41.702: INFO @log_profile : T read data: 1.433693 2019-10-09 05:42:41.702: INFO @log_profile : T hooks: 3.471314 2019-10-09 05:42:41.702: INFO @main_loop : Epoch 94 done 2019-10-09 05:42:41.702: INFO @main_loop : Training epoch 95 2019-10-09 05:45:50.138: INFO @log_variables: train loss mean: 0.279547 2019-10-09 05:45:50.138: INFO @log_variables: train age_loss mean: 4.653752 2019-10-09 05:45:50.138: INFO @log_variables: train gender_loss mean: 0.051619 2019-10-09 05:45:50.138: INFO @log_variables: train matching_loss nanmean: 0.349602 2019-10-09 05:45:50.138: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:45:50.138: INFO @log_variables: train age_mae mean: 5.130797 2019-10-09 05:45:50.138: INFO @log_variables: train gender_accuracy mean: 0.981277 2019-10-09 05:45:50.138: INFO @log_variables: train positive_distance nanmean: 0.758154 2019-10-09 05:45:50.138: INFO @log_variables: train negative_distance nanmean: 1.409118 2019-10-09 05:45:50.138: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:45:50.138: INFO @log_variables: valid loss mean: 0.453073 2019-10-09 05:45:50.138: INFO @log_variables: valid age_loss mean: 6.418425 2019-10-09 05:45:50.138: INFO @log_variables: valid gender_loss mean: 0.259686 2019-10-09 05:45:50.139: INFO @log_variables: valid matching_loss nanmean: 0.502997 2019-10-09 05:45:50.139: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:45:50.139: INFO @log_variables: valid age_mae mean: 6.901840 2019-10-09 05:45:50.139: INFO @log_variables: valid gender_accuracy mean: 0.918855 2019-10-09 05:45:50.139: INFO @log_variables: valid positive_distance nanmean: 0.784154 2019-10-09 05:45:50.139: INFO @log_variables: valid negative_distance nanmean: 1.372973 2019-10-09 05:45:50.139: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:45:50.139: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:45:52.360: INFO @metrics_hook: valid matching accuracy: 0.8887689632427894, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:45:52.818: INFO @decay_lr : LR updated to `0.0006211446` 2019-10-09 05:45:52.820: INFO @log_profile : T train: 178.023004 2019-10-09 05:45:52.820: INFO @log_profile : T valid: 8.311258 2019-10-09 05:45:52.820: INFO @log_profile : T read data: 1.451026 2019-10-09 05:45:52.820: INFO @log_profile : T hooks: 3.245773 2019-10-09 05:45:52.820: INFO @main_loop : Epoch 95 done 2019-10-09 05:45:52.820: INFO @main_loop : Training epoch 96 2019-10-09 05:49:01.123: INFO @log_variables: train loss mean: 0.279374 2019-10-09 05:49:01.123: INFO @log_variables: train age_loss mean: 4.676937 2019-10-09 05:49:01.123: INFO @log_variables: train gender_loss mean: 0.050707 2019-10-09 05:49:01.123: INFO @log_variables: train matching_loss nanmean: 0.347660 2019-10-09 05:49:01.123: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:49:01.123: INFO @log_variables: train age_mae mean: 5.154311 2019-10-09 05:49:01.123: INFO @log_variables: train gender_accuracy mean: 0.982025 2019-10-09 05:49:01.123: INFO @log_variables: train positive_distance nanmean: 0.757596 2019-10-09 05:49:01.124: INFO @log_variables: train negative_distance nanmean: 1.409131 2019-10-09 05:49:01.124: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:49:01.124: INFO @log_variables: valid loss mean: 0.459077 2019-10-09 05:49:01.124: INFO @log_variables: valid age_loss mean: 6.582639 2019-10-09 05:49:01.124: INFO @log_variables: valid gender_loss mean: 0.263444 2019-10-09 05:49:01.124: INFO @log_variables: valid matching_loss nanmean: 0.501431 2019-10-09 05:49:01.124: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:49:01.124: INFO @log_variables: valid age_mae mean: 7.067312 2019-10-09 05:49:01.124: INFO @log_variables: valid gender_accuracy mean: 0.920274 2019-10-09 05:49:01.124: INFO @log_variables: valid positive_distance nanmean: 0.786113 2019-10-09 05:49:01.124: INFO @log_variables: valid negative_distance nanmean: 1.373335 2019-10-09 05:49:01.124: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:49:01.124: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:49:03.477: INFO @metrics_hook: valid matching accuracy: 0.8864903759669005, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:49:03.889: INFO @decay_lr : LR updated to `0.0006180389` 2019-10-09 05:49:03.891: INFO @log_profile : T train: 178.292271 2019-10-09 05:49:03.891: INFO @log_profile : T valid: 8.342200 2019-10-09 05:49:03.891: INFO @log_profile : T read data: 1.016653 2019-10-09 05:49:03.891: INFO @log_profile : T hooks: 3.333464 2019-10-09 05:49:03.891: INFO @main_loop : Epoch 96 done 2019-10-09 05:49:03.891: INFO @main_loop : Training epoch 97 2019-10-09 05:52:12.425: INFO @log_variables: train loss mean: 0.279991 2019-10-09 05:52:12.425: INFO @log_variables: train age_loss mean: 4.669596 2019-10-09 05:52:12.425: INFO @log_variables: train gender_loss mean: 0.052451 2019-10-09 05:52:12.425: INFO @log_variables: train matching_loss nanmean: 0.348562 2019-10-09 05:52:12.425: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:52:12.425: INFO @log_variables: train age_mae mean: 5.147172 2019-10-09 05:52:12.425: INFO @log_variables: train gender_accuracy mean: 0.981825 2019-10-09 05:52:12.425: INFO @log_variables: train positive_distance nanmean: 0.758607 2019-10-09 05:52:12.426: INFO @log_variables: train negative_distance nanmean: 1.409028 2019-10-09 05:52:12.426: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:52:12.426: INFO @log_variables: valid loss mean: 0.444961 2019-10-09 05:52:12.426: INFO @log_variables: valid age_loss mean: 6.310111 2019-10-09 05:52:12.426: INFO @log_variables: valid gender_loss mean: 0.248548 2019-10-09 05:52:12.426: INFO @log_variables: valid matching_loss nanmean: 0.499819 2019-10-09 05:52:12.426: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:52:12.426: INFO @log_variables: valid age_mae mean: 6.792102 2019-10-09 05:52:12.426: INFO @log_variables: valid gender_accuracy mean: 0.923172 2019-10-09 05:52:12.426: INFO @log_variables: valid positive_distance nanmean: 0.790154 2019-10-09 05:52:12.426: INFO @log_variables: valid negative_distance nanmean: 1.375662 2019-10-09 05:52:12.426: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:52:12.426: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:52:14.669: INFO @metrics_hook: valid matching accuracy: 0.8881093721892427, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:52:15.063: INFO @decay_lr : LR updated to `0.0006149487` 2019-10-09 05:52:15.064: INFO @log_profile : T train: 178.162599 2019-10-09 05:52:15.064: INFO @log_profile : T valid: 8.312740 2019-10-09 05:52:15.065: INFO @log_profile : T read data: 1.426188 2019-10-09 05:52:15.065: INFO @log_profile : T hooks: 3.185759 2019-10-09 05:52:15.065: INFO @main_loop : Epoch 97 done 2019-10-09 05:52:15.065: INFO @main_loop : Training epoch 98 2019-10-09 05:55:23.831: INFO @log_variables: train loss mean: 0.278911 2019-10-09 05:55:23.832: INFO @log_variables: train age_loss mean: 4.644335 2019-10-09 05:55:23.832: INFO @log_variables: train gender_loss mean: 0.052967 2019-10-09 05:55:23.832: INFO @log_variables: train matching_loss nanmean: 0.347225 2019-10-09 05:55:23.832: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:55:23.832: INFO @log_variables: train age_mae mean: 5.122154 2019-10-09 05:55:23.832: INFO @log_variables: train gender_accuracy mean: 0.980708 2019-10-09 05:55:23.832: INFO @log_variables: train positive_distance nanmean: 0.758433 2019-10-09 05:55:23.832: INFO @log_variables: train negative_distance nanmean: 1.409232 2019-10-09 05:55:23.832: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:55:23.832: INFO @log_variables: valid loss mean: 0.455880 2019-10-09 05:55:23.832: INFO @log_variables: valid age_loss mean: 6.452770 2019-10-09 05:55:23.832: INFO @log_variables: valid gender_loss mean: 0.261411 2019-10-09 05:55:23.832: INFO @log_variables: valid matching_loss nanmean: 0.506541 2019-10-09 05:55:23.832: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:55:23.832: INFO @log_variables: valid age_mae mean: 6.936298 2019-10-09 05:55:23.832: INFO @log_variables: valid gender_accuracy mean: 0.919742 2019-10-09 05:55:23.832: INFO @log_variables: valid positive_distance nanmean: 0.790702 2019-10-09 05:55:23.832: INFO @log_variables: valid negative_distance nanmean: 1.371588 2019-10-09 05:55:23.832: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:55:23.833: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:55:25.924: INFO @metrics_hook: valid matching accuracy: 0.8878095580739941, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:55:26.320: INFO @decay_lr : LR updated to `0.0006118739` 2019-10-09 05:55:26.321: INFO @log_profile : T train: 178.310081 2019-10-09 05:55:26.321: INFO @log_profile : T valid: 8.309545 2019-10-09 05:55:26.321: INFO @log_profile : T read data: 1.465427 2019-10-09 05:55:26.321: INFO @log_profile : T hooks: 3.084003 2019-10-09 05:55:26.321: INFO @main_loop : Epoch 98 done 2019-10-09 05:55:26.322: INFO @main_loop : Training epoch 99 2019-10-09 05:58:34.543: INFO @log_variables: train loss mean: 0.275734 2019-10-09 05:58:34.543: INFO @log_variables: train age_loss mean: 4.595353 2019-10-09 05:58:34.543: INFO @log_variables: train gender_loss mean: 0.049043 2019-10-09 05:58:34.543: INFO @log_variables: train matching_loss nanmean: 0.346195 2019-10-09 05:58:34.543: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 05:58:34.543: INFO @log_variables: train age_mae mean: 5.072264 2019-10-09 05:58:34.543: INFO @log_variables: train gender_accuracy mean: 0.982044 2019-10-09 05:58:34.543: INFO @log_variables: train positive_distance nanmean: 0.757776 2019-10-09 05:58:34.543: INFO @log_variables: train negative_distance nanmean: 1.408841 2019-10-09 05:58:34.543: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 05:58:34.543: INFO @log_variables: valid loss mean: 0.451062 2019-10-09 05:58:34.543: INFO @log_variables: valid age_loss mean: 6.431488 2019-10-09 05:58:34.543: INFO @log_variables: valid gender_loss mean: 0.251362 2019-10-09 05:58:34.544: INFO @log_variables: valid matching_loss nanmean: 0.503782 2019-10-09 05:58:34.544: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 05:58:34.544: INFO @log_variables: valid age_mae mean: 6.914265 2019-10-09 05:58:34.544: INFO @log_variables: valid gender_accuracy mean: 0.919565 2019-10-09 05:58:34.544: INFO @log_variables: valid positive_distance nanmean: 0.785729 2019-10-09 05:58:34.544: INFO @log_variables: valid negative_distance nanmean: 1.370465 2019-10-09 05:58:34.544: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 05:58:34.544: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 05:58:36.522: INFO @metrics_hook: valid matching accuracy: 0.889068777358038, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 05:58:36.938: INFO @decay_lr : LR updated to `0.00060881453` 2019-10-09 05:58:36.939: INFO @log_profile : T train: 178.127991 2019-10-09 05:58:36.939: INFO @log_profile : T valid: 8.420509 2019-10-09 05:58:36.939: INFO @log_profile : T read data: 1.005365 2019-10-09 05:58:36.939: INFO @log_profile : T hooks: 2.978363 2019-10-09 05:58:36.939: INFO @main_loop : Epoch 99 done 2019-10-09 05:58:36.939: INFO @main_loop : Training epoch 100 2019-10-09 06:01:45.795: INFO @log_variables: train loss mean: 0.276147 2019-10-09 06:01:45.796: INFO @log_variables: train age_loss mean: 4.601223 2019-10-09 06:01:45.796: INFO @log_variables: train gender_loss mean: 0.049056 2019-10-09 06:01:45.796: INFO @log_variables: train matching_loss nanmean: 0.346877 2019-10-09 06:01:45.796: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:01:45.796: INFO @log_variables: train age_mae mean: 5.078331 2019-10-09 06:01:45.796: INFO @log_variables: train gender_accuracy mean: 0.981981 2019-10-09 06:01:45.796: INFO @log_variables: train positive_distance nanmean: 0.757148 2019-10-09 06:01:45.796: INFO @log_variables: train negative_distance nanmean: 1.408977 2019-10-09 06:01:45.796: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:01:45.796: INFO @log_variables: valid loss mean: 0.458891 2019-10-09 06:01:45.796: INFO @log_variables: valid age_loss mean: 6.553871 2019-10-09 06:01:45.796: INFO @log_variables: valid gender_loss mean: 0.261598 2019-10-09 06:01:45.796: INFO @log_variables: valid matching_loss nanmean: 0.505578 2019-10-09 06:01:45.796: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:01:45.796: INFO @log_variables: valid age_mae mean: 7.037896 2019-10-09 06:01:45.796: INFO @log_variables: valid gender_accuracy mean: 0.920925 2019-10-09 06:01:45.796: INFO @log_variables: valid positive_distance nanmean: 0.782815 2019-10-09 06:01:45.796: INFO @log_variables: valid negative_distance nanmean: 1.368952 2019-10-09 06:01:45.796: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:01:45.797: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:01:47.913: INFO @metrics_hook: valid matching accuracy: 0.8880494093661929, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:01:47.963: INFO @metrics_hook: valid age_mae: 7.038 +-0.094 (16908) 2019-10-09 06:01:47.964: INFO @metrics_hook: valid gender_accuracy: 0.921 +-0.004 (16908) 2019-10-09 06:01:49.864: INFO @decay_lr : LR updated to `0.00060577044` 2019-10-09 06:01:49.865: INFO @log_profile : T train: 178.386336 2019-10-09 06:01:49.865: INFO @log_profile : T valid: 8.338349 2019-10-09 06:01:49.865: INFO @log_profile : T read data: 1.472528 2019-10-09 06:01:49.865: INFO @log_profile : T hooks: 4.642926 2019-10-09 06:01:49.865: INFO @main_loop : Epoch 100 done 2019-10-09 06:01:49.865: INFO @main_loop : Training epoch 101 2019-10-09 06:05:00.021: INFO @log_variables: train loss mean: 0.277302 2019-10-09 06:05:00.021: INFO @log_variables: train age_loss mean: 4.645889 2019-10-09 06:05:00.021: INFO @log_variables: train gender_loss mean: 0.050115 2019-10-09 06:05:00.021: INFO @log_variables: train matching_loss nanmean: 0.344931 2019-10-09 06:05:00.022: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:05:00.022: INFO @log_variables: train age_mae mean: 5.123201 2019-10-09 06:05:00.022: INFO @log_variables: train gender_accuracy mean: 0.982156 2019-10-09 06:05:00.022: INFO @log_variables: train positive_distance nanmean: 0.755909 2019-10-09 06:05:00.022: INFO @log_variables: train negative_distance nanmean: 1.409105 2019-10-09 06:05:00.022: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:05:00.022: INFO @log_variables: valid loss mean: 0.449980 2019-10-09 06:05:00.022: INFO @log_variables: valid age_loss mean: 6.540785 2019-10-09 06:05:00.022: INFO @log_variables: valid gender_loss mean: 0.238508 2019-10-09 06:05:00.022: INFO @log_variables: valid matching_loss nanmean: 0.502351 2019-10-09 06:05:00.022: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:05:00.022: INFO @log_variables: valid age_mae mean: 7.022996 2019-10-09 06:05:00.022: INFO @log_variables: valid gender_accuracy mean: 0.924119 2019-10-09 06:05:00.022: INFO @log_variables: valid positive_distance nanmean: 0.788016 2019-10-09 06:05:00.022: INFO @log_variables: valid negative_distance nanmean: 1.375866 2019-10-09 06:05:00.022: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:05:00.023: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:05:02.557: INFO @metrics_hook: valid matching accuracy: 0.8866103016129999, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:05:02.993: INFO @decay_lr : LR updated to `0.0006027416` 2019-10-09 06:05:02.994: INFO @log_profile : T train: 179.945631 2019-10-09 06:05:02.994: INFO @log_profile : T valid: 8.547585 2019-10-09 06:05:02.994: INFO @log_profile : T read data: 0.988249 2019-10-09 06:05:02.994: INFO @log_profile : T hooks: 3.560022 2019-10-09 06:05:02.994: INFO @main_loop : Epoch 101 done 2019-10-09 06:05:02.994: INFO @main_loop : Training epoch 102 2019-10-09 06:08:13.807: INFO @log_variables: train loss mean: 0.276281 2019-10-09 06:08:13.807: INFO @log_variables: train age_loss mean: 4.634440 2019-10-09 06:08:13.807: INFO @log_variables: train gender_loss mean: 0.048159 2019-10-09 06:08:13.807: INFO @log_variables: train matching_loss nanmean: 0.344868 2019-10-09 06:08:13.808: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:08:13.808: INFO @log_variables: train age_mae mean: 5.111901 2019-10-09 06:08:13.808: INFO @log_variables: train gender_accuracy mean: 0.982560 2019-10-09 06:08:13.808: INFO @log_variables: train positive_distance nanmean: 0.756014 2019-10-09 06:08:13.808: INFO @log_variables: train negative_distance nanmean: 1.408775 2019-10-09 06:08:13.808: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:08:13.808: INFO @log_variables: valid loss mean: 0.460897 2019-10-09 06:08:13.808: INFO @log_variables: valid age_loss mean: 6.589169 2019-10-09 06:08:13.808: INFO @log_variables: valid gender_loss mean: 0.265040 2019-10-09 06:08:13.808: INFO @log_variables: valid matching_loss nanmean: 0.504825 2019-10-09 06:08:13.808: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:08:13.808: INFO @log_variables: valid age_mae mean: 7.073748 2019-10-09 06:08:13.808: INFO @log_variables: valid gender_accuracy mean: 0.921221 2019-10-09 06:08:13.808: INFO @log_variables: valid positive_distance nanmean: 0.785773 2019-10-09 06:08:13.808: INFO @log_variables: valid negative_distance nanmean: 1.373240 2019-10-09 06:08:13.808: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:08:13.808: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:08:16.009: INFO @metrics_hook: valid matching accuracy: 0.8889488517119386, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:08:16.442: INFO @decay_lr : LR updated to `0.0005997279` 2019-10-09 06:08:16.444: INFO @log_profile : T train: 180.167976 2019-10-09 06:08:16.444: INFO @log_profile : T valid: 8.516901 2019-10-09 06:08:16.444: INFO @log_profile : T read data: 1.430657 2019-10-09 06:08:16.444: INFO @log_profile : T hooks: 3.249445 2019-10-09 06:08:16.444: INFO @main_loop : Epoch 102 done 2019-10-09 06:08:16.444: INFO @main_loop : Training epoch 103 2019-10-09 06:11:27.066: INFO @log_variables: train loss mean: 0.276410 2019-10-09 06:11:27.067: INFO @log_variables: train age_loss mean: 4.629822 2019-10-09 06:11:27.067: INFO @log_variables: train gender_loss mean: 0.050232 2019-10-09 06:11:27.067: INFO @log_variables: train matching_loss nanmean: 0.343658 2019-10-09 06:11:27.067: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:11:27.067: INFO @log_variables: train age_mae mean: 5.106997 2019-10-09 06:11:27.067: INFO @log_variables: train gender_accuracy mean: 0.981935 2019-10-09 06:11:27.067: INFO @log_variables: train positive_distance nanmean: 0.756098 2019-10-09 06:11:27.067: INFO @log_variables: train negative_distance nanmean: 1.408877 2019-10-09 06:11:27.067: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:11:27.067: INFO @log_variables: valid loss mean: 0.445685 2019-10-09 06:11:27.067: INFO @log_variables: valid age_loss mean: 6.276963 2019-10-09 06:11:27.067: INFO @log_variables: valid gender_loss mean: 0.250543 2019-10-09 06:11:27.067: INFO @log_variables: valid matching_loss nanmean: 0.503383 2019-10-09 06:11:27.067: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:11:27.067: INFO @log_variables: valid age_mae mean: 6.759355 2019-10-09 06:11:27.067: INFO @log_variables: valid gender_accuracy mean: 0.920274 2019-10-09 06:11:27.067: INFO @log_variables: valid positive_distance nanmean: 0.782523 2019-10-09 06:11:27.067: INFO @log_variables: valid negative_distance nanmean: 1.372645 2019-10-09 06:11:27.067: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:11:27.068: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:11:29.164: INFO @metrics_hook: valid matching accuracy: 0.8887689632427894, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:11:29.627: INFO @decay_lr : LR updated to `0.0005967293` 2019-10-09 06:11:29.629: INFO @log_profile : T train: 180.070437 2019-10-09 06:11:29.629: INFO @log_profile : T valid: 8.466146 2019-10-09 06:11:29.629: INFO @log_profile : T read data: 1.394080 2019-10-09 06:11:29.629: INFO @log_profile : T hooks: 3.169716 2019-10-09 06:11:29.629: INFO @main_loop : Epoch 103 done 2019-10-09 06:11:29.629: INFO @main_loop : Training epoch 104 2019-10-09 06:14:39.793: INFO @log_variables: train loss mean: 0.275202 2019-10-09 06:14:39.793: INFO @log_variables: train age_loss mean: 4.594884 2019-10-09 06:14:39.793: INFO @log_variables: train gender_loss mean: 0.047628 2019-10-09 06:14:39.793: INFO @log_variables: train matching_loss nanmean: 0.346008 2019-10-09 06:14:39.793: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:14:39.793: INFO @log_variables: train age_mae mean: 5.071589 2019-10-09 06:14:39.793: INFO @log_variables: train gender_accuracy mean: 0.982964 2019-10-09 06:14:39.793: INFO @log_variables: train positive_distance nanmean: 0.758028 2019-10-09 06:14:39.793: INFO @log_variables: train negative_distance nanmean: 1.409216 2019-10-09 06:14:39.793: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:14:39.793: INFO @log_variables: valid loss mean: 0.457500 2019-10-09 06:14:39.793: INFO @log_variables: valid age_loss mean: 6.397094 2019-10-09 06:14:39.793: INFO @log_variables: valid gender_loss mean: 0.272893 2019-10-09 06:14:39.793: INFO @log_variables: valid matching_loss nanmean: 0.505647 2019-10-09 06:14:39.793: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:14:39.793: INFO @log_variables: valid age_mae mean: 6.880581 2019-10-09 06:14:39.794: INFO @log_variables: valid gender_accuracy mean: 0.917731 2019-10-09 06:14:39.794: INFO @log_variables: valid positive_distance nanmean: 0.775138 2019-10-09 06:14:39.794: INFO @log_variables: valid negative_distance nanmean: 1.367856 2019-10-09 06:14:39.794: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:14:39.794: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:14:41.935: INFO @metrics_hook: valid matching accuracy: 0.8884691491275409, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:14:42.361: INFO @decay_lr : LR updated to `0.0005937456` 2019-10-09 06:14:42.362: INFO @log_profile : T train: 179.914018 2019-10-09 06:14:42.363: INFO @log_profile : T valid: 8.542254 2019-10-09 06:14:42.363: INFO @log_profile : T read data: 1.026730 2019-10-09 06:14:42.363: INFO @log_profile : T hooks: 3.163763 2019-10-09 06:14:42.363: INFO @main_loop : Epoch 104 done 2019-10-09 06:14:42.363: INFO @main_loop : Training epoch 105 2019-10-09 06:17:52.956: INFO @log_variables: train loss mean: 0.272306 2019-10-09 06:17:52.956: INFO @log_variables: train age_loss mean: 4.555086 2019-10-09 06:17:52.956: INFO @log_variables: train gender_loss mean: 0.045526 2019-10-09 06:17:52.956: INFO @log_variables: train matching_loss nanmean: 0.343114 2019-10-09 06:17:52.956: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:17:52.956: INFO @log_variables: train age_mae mean: 5.032053 2019-10-09 06:17:52.956: INFO @log_variables: train gender_accuracy mean: 0.983347 2019-10-09 06:17:52.956: INFO @log_variables: train positive_distance nanmean: 0.756152 2019-10-09 06:17:52.956: INFO @log_variables: train negative_distance nanmean: 1.408889 2019-10-09 06:17:52.956: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:17:52.956: INFO @log_variables: valid loss mean: 0.461663 2019-10-09 06:17:52.956: INFO @log_variables: valid age_loss mean: 6.375747 2019-10-09 06:17:52.956: INFO @log_variables: valid gender_loss mean: 0.294582 2019-10-09 06:17:52.956: INFO @log_variables: valid matching_loss nanmean: 0.498999 2019-10-09 06:17:52.956: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:17:52.957: INFO @log_variables: valid age_mae mean: 6.858969 2019-10-09 06:17:52.957: INFO @log_variables: valid gender_accuracy mean: 0.917968 2019-10-09 06:17:52.957: INFO @log_variables: valid positive_distance nanmean: 0.788841 2019-10-09 06:17:52.957: INFO @log_variables: valid negative_distance nanmean: 1.374940 2019-10-09 06:17:52.957: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:17:52.957: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:17:55.265: INFO @metrics_hook: valid matching accuracy: 0.8884091863044912, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:17:55.701: INFO @decay_lr : LR updated to `0.0005907769` 2019-10-09 06:17:55.702: INFO @log_profile : T train: 179.860739 2019-10-09 06:17:55.702: INFO @log_profile : T valid: 8.515367 2019-10-09 06:17:55.702: INFO @log_profile : T read data: 1.558491 2019-10-09 06:17:55.702: INFO @log_profile : T hooks: 3.319634 2019-10-09 06:17:55.702: INFO @main_loop : Epoch 105 done 2019-10-09 06:17:55.702: INFO @main_loop : Training epoch 106 2019-10-09 06:21:05.980: INFO @log_variables: train loss mean: 0.272469 2019-10-09 06:21:05.980: INFO @log_variables: train age_loss mean: 4.558065 2019-10-09 06:21:05.980: INFO @log_variables: train gender_loss mean: 0.047299 2019-10-09 06:21:05.980: INFO @log_variables: train matching_loss nanmean: 0.341549 2019-10-09 06:21:05.980: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:21:05.980: INFO @log_variables: train age_mae mean: 5.034612 2019-10-09 06:21:05.980: INFO @log_variables: train gender_accuracy mean: 0.983050 2019-10-09 06:21:05.980: INFO @log_variables: train positive_distance nanmean: 0.754639 2019-10-09 06:21:05.980: INFO @log_variables: train negative_distance nanmean: 1.409101 2019-10-09 06:21:05.980: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:21:05.980: INFO @log_variables: valid loss mean: 0.457326 2019-10-09 06:21:05.980: INFO @log_variables: valid age_loss mean: 6.438622 2019-10-09 06:21:05.980: INFO @log_variables: valid gender_loss mean: 0.267888 2019-10-09 06:21:05.980: INFO @log_variables: valid matching_loss nanmean: 0.505959 2019-10-09 06:21:05.980: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:21:05.981: INFO @log_variables: valid age_mae mean: 6.922002 2019-10-09 06:21:05.981: INFO @log_variables: valid gender_accuracy mean: 0.921162 2019-10-09 06:21:05.981: INFO @log_variables: valid positive_distance nanmean: 0.787071 2019-10-09 06:21:05.981: INFO @log_variables: valid negative_distance nanmean: 1.373327 2019-10-09 06:21:05.981: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:21:05.981: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:21:08.714: INFO @metrics_hook: valid matching accuracy: 0.8870300413743479, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:21:09.168: INFO @decay_lr : LR updated to `0.00058782304` 2019-10-09 06:21:09.169: INFO @log_profile : T train: 179.996190 2019-10-09 06:21:09.169: INFO @log_profile : T valid: 8.562280 2019-10-09 06:21:09.169: INFO @log_profile : T read data: 1.019560 2019-10-09 06:21:09.169: INFO @log_profile : T hooks: 3.802272 2019-10-09 06:21:09.169: INFO @main_loop : Epoch 106 done 2019-10-09 06:21:09.169: INFO @main_loop : Training epoch 107 2019-10-09 06:24:20.135: INFO @log_variables: train loss mean: 0.272144 2019-10-09 06:24:20.135: INFO @log_variables: train age_loss mean: 4.532963 2019-10-09 06:24:20.135: INFO @log_variables: train gender_loss mean: 0.047497 2019-10-09 06:24:20.135: INFO @log_variables: train matching_loss nanmean: 0.342852 2019-10-09 06:24:20.135: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:24:20.135: INFO @log_variables: train age_mae mean: 5.009643 2019-10-09 06:24:20.135: INFO @log_variables: train gender_accuracy mean: 0.983076 2019-10-09 06:24:20.135: INFO @log_variables: train positive_distance nanmean: 0.755665 2019-10-09 06:24:20.135: INFO @log_variables: train negative_distance nanmean: 1.409067 2019-10-09 06:24:20.135: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:24:20.135: INFO @log_variables: valid loss mean: 0.451441 2019-10-09 06:24:20.135: INFO @log_variables: valid age_loss mean: 6.393328 2019-10-09 06:24:20.135: INFO @log_variables: valid gender_loss mean: 0.254853 2019-10-09 06:24:20.135: INFO @log_variables: valid matching_loss nanmean: 0.505282 2019-10-09 06:24:20.136: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:24:20.136: INFO @log_variables: valid age_mae mean: 6.876211 2019-10-09 06:24:20.136: INFO @log_variables: valid gender_accuracy mean: 0.924060 2019-10-09 06:24:20.136: INFO @log_variables: valid positive_distance nanmean: 0.787277 2019-10-09 06:24:20.136: INFO @log_variables: valid negative_distance nanmean: 1.372528 2019-10-09 06:24:20.136: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:24:20.136: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:24:22.379: INFO @metrics_hook: valid matching accuracy: 0.8875697067817953, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:24:22.815: INFO @decay_lr : LR updated to `0.0005848839` 2019-10-09 06:24:22.816: INFO @log_profile : T train: 180.006617 2019-10-09 06:24:22.816: INFO @log_profile : T valid: 8.502334 2019-10-09 06:24:22.816: INFO @log_profile : T read data: 1.772388 2019-10-09 06:24:22.816: INFO @log_profile : T hooks: 3.279200 2019-10-09 06:24:22.816: INFO @main_loop : Epoch 107 done 2019-10-09 06:24:22.816: INFO @main_loop : Training epoch 108 2019-10-09 06:27:33.453: INFO @log_variables: train loss mean: 0.272807 2019-10-09 06:27:33.453: INFO @log_variables: train age_loss mean: 4.580779 2019-10-09 06:27:33.453: INFO @log_variables: train gender_loss mean: 0.046564 2019-10-09 06:27:33.453: INFO @log_variables: train matching_loss nanmean: 0.341062 2019-10-09 06:27:33.453: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:27:33.453: INFO @log_variables: train age_mae mean: 5.057538 2019-10-09 06:27:33.454: INFO @log_variables: train gender_accuracy mean: 0.983123 2019-10-09 06:27:33.454: INFO @log_variables: train positive_distance nanmean: 0.755153 2019-10-09 06:27:33.454: INFO @log_variables: train negative_distance nanmean: 1.409029 2019-10-09 06:27:33.454: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:27:33.454: INFO @log_variables: valid loss mean: 0.447043 2019-10-09 06:27:33.454: INFO @log_variables: valid age_loss mean: 6.360890 2019-10-09 06:27:33.454: INFO @log_variables: valid gender_loss mean: 0.241909 2019-10-09 06:27:33.454: INFO @log_variables: valid matching_loss nanmean: 0.507835 2019-10-09 06:27:33.454: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:27:33.454: INFO @log_variables: valid age_mae mean: 6.843613 2019-10-09 06:27:33.454: INFO @log_variables: valid gender_accuracy mean: 0.924651 2019-10-09 06:27:33.454: INFO @log_variables: valid positive_distance nanmean: 0.785989 2019-10-09 06:27:33.454: INFO @log_variables: valid negative_distance nanmean: 1.370933 2019-10-09 06:27:33.454: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:27:33.454: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:27:35.684: INFO @metrics_hook: valid matching accuracy: 0.8876896324278947, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:27:36.134: INFO @decay_lr : LR updated to `0.0005819595` 2019-10-09 06:27:36.135: INFO @log_profile : T train: 179.951413 2019-10-09 06:27:36.135: INFO @log_profile : T valid: 8.534262 2019-10-09 06:27:36.135: INFO @log_profile : T read data: 1.458123 2019-10-09 06:27:36.135: INFO @log_profile : T hooks: 3.287409 2019-10-09 06:27:36.135: INFO @main_loop : Epoch 108 done 2019-10-09 06:27:36.135: INFO @main_loop : Training epoch 109 2019-10-09 06:30:46.264: INFO @log_variables: train loss mean: 0.269972 2019-10-09 06:30:46.264: INFO @log_variables: train age_loss mean: 4.506560 2019-10-09 06:30:46.264: INFO @log_variables: train gender_loss mean: 0.045736 2019-10-09 06:30:46.264: INFO @log_variables: train matching_loss nanmean: 0.340522 2019-10-09 06:30:46.264: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:30:46.264: INFO @log_variables: train age_mae mean: 4.982874 2019-10-09 06:30:46.264: INFO @log_variables: train gender_accuracy mean: 0.983546 2019-10-09 06:30:46.264: INFO @log_variables: train positive_distance nanmean: 0.753908 2019-10-09 06:30:46.265: INFO @log_variables: train negative_distance nanmean: 1.408723 2019-10-09 06:30:46.265: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:30:46.265: INFO @log_variables: valid loss mean: 0.463696 2019-10-09 06:30:46.265: INFO @log_variables: valid age_loss mean: 6.483588 2019-10-09 06:30:46.265: INFO @log_variables: valid gender_loss mean: 0.284494 2019-10-09 06:30:46.265: INFO @log_variables: valid matching_loss nanmean: 0.504604 2019-10-09 06:30:46.265: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:30:46.265: INFO @log_variables: valid age_mae mean: 6.967241 2019-10-09 06:30:46.265: INFO @log_variables: valid gender_accuracy mean: 0.913236 2019-10-09 06:30:46.265: INFO @log_variables: valid positive_distance nanmean: 0.785734 2019-10-09 06:30:46.265: INFO @log_variables: valid negative_distance nanmean: 1.372063 2019-10-09 06:30:46.265: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:30:46.265: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:30:48.752: INFO @metrics_hook: valid matching accuracy: 0.8867302272590993, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:30:49.198: INFO @decay_lr : LR updated to `0.0005790497` 2019-10-09 06:30:49.199: INFO @log_profile : T train: 179.895487 2019-10-09 06:30:49.199: INFO @log_profile : T valid: 8.533299 2019-10-09 06:30:49.199: INFO @log_profile : T read data: 1.032127 2019-10-09 06:30:49.199: INFO @log_profile : T hooks: 3.516009 2019-10-09 06:30:49.199: INFO @main_loop : Epoch 109 done 2019-10-09 06:30:49.200: INFO @main_loop : Training epoch 110 2019-10-09 06:33:59.870: INFO @log_variables: train loss mean: 0.270875 2019-10-09 06:33:59.870: INFO @log_variables: train age_loss mean: 4.543294 2019-10-09 06:33:59.871: INFO @log_variables: train gender_loss mean: 0.044956 2019-10-09 06:33:59.871: INFO @log_variables: train matching_loss nanmean: 0.340427 2019-10-09 06:33:59.871: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:33:59.871: INFO @log_variables: train age_mae mean: 5.020097 2019-10-09 06:33:59.871: INFO @log_variables: train gender_accuracy mean: 0.984061 2019-10-09 06:33:59.871: INFO @log_variables: train positive_distance nanmean: 0.754749 2019-10-09 06:33:59.871: INFO @log_variables: train negative_distance nanmean: 1.408921 2019-10-09 06:33:59.871: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:33:59.871: INFO @log_variables: valid loss mean: 0.464771 2019-10-09 06:33:59.871: INFO @log_variables: valid age_loss mean: 6.529121 2019-10-09 06:33:59.871: INFO @log_variables: valid gender_loss mean: 0.282776 2019-10-09 06:33:59.871: INFO @log_variables: valid matching_loss nanmean: 0.505102 2019-10-09 06:33:59.871: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:33:59.871: INFO @log_variables: valid age_mae mean: 7.013316 2019-10-09 06:33:59.871: INFO @log_variables: valid gender_accuracy mean: 0.918145 2019-10-09 06:33:59.871: INFO @log_variables: valid positive_distance nanmean: 0.785050 2019-10-09 06:33:59.872: INFO @log_variables: valid negative_distance nanmean: 1.374620 2019-10-09 06:33:59.872: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:33:59.872: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:34:01.985: INFO @metrics_hook: valid matching accuracy: 0.889068777358038, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:34:02.429: INFO @decay_lr : LR updated to `0.00057615445` 2019-10-09 06:34:02.430: INFO @log_profile : T train: 179.956704 2019-10-09 06:34:02.430: INFO @log_profile : T valid: 8.538980 2019-10-09 06:34:02.430: INFO @log_profile : T read data: 1.473162 2019-10-09 06:34:02.430: INFO @log_profile : T hooks: 3.176495 2019-10-09 06:34:02.430: INFO @main_loop : Epoch 110 done 2019-10-09 06:34:02.430: INFO @main_loop : Training epoch 111 2019-10-09 06:37:13.060: INFO @log_variables: train loss mean: 0.272488 2019-10-09 06:37:13.060: INFO @log_variables: train age_loss mean: 4.540032 2019-10-09 06:37:13.060: INFO @log_variables: train gender_loss mean: 0.048754 2019-10-09 06:37:13.060: INFO @log_variables: train matching_loss nanmean: 0.341956 2019-10-09 06:37:13.060: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:37:13.061: INFO @log_variables: train age_mae mean: 5.016912 2019-10-09 06:37:13.061: INFO @log_variables: train gender_accuracy mean: 0.982805 2019-10-09 06:37:13.061: INFO @log_variables: train positive_distance nanmean: 0.754844 2019-10-09 06:37:13.061: INFO @log_variables: train negative_distance nanmean: 1.408736 2019-10-09 06:37:13.061: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:37:13.061: INFO @log_variables: valid loss mean: 0.453696 2019-10-09 06:37:13.061: INFO @log_variables: valid age_loss mean: 6.338889 2019-10-09 06:37:13.061: INFO @log_variables: valid gender_loss mean: 0.272586 2019-10-09 06:37:13.061: INFO @log_variables: valid matching_loss nanmean: 0.499982 2019-10-09 06:37:13.061: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:37:13.061: INFO @log_variables: valid age_mae mean: 6.821850 2019-10-09 06:37:13.061: INFO @log_variables: valid gender_accuracy mean: 0.919092 2019-10-09 06:37:13.061: INFO @log_variables: valid positive_distance nanmean: 0.781807 2019-10-09 06:37:13.061: INFO @log_variables: valid negative_distance nanmean: 1.375523 2019-10-09 06:37:13.061: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:37:13.061: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:37:15.264: INFO @metrics_hook: valid matching accuracy: 0.8888289260658392, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:37:15.716: INFO @decay_lr : LR updated to `0.0005732737` 2019-10-09 06:37:15.717: INFO @log_profile : T train: 179.937762 2019-10-09 06:37:15.718: INFO @log_profile : T valid: 8.544444 2019-10-09 06:37:15.718: INFO @log_profile : T read data: 1.477468 2019-10-09 06:37:15.718: INFO @log_profile : T hooks: 3.240930 2019-10-09 06:37:15.718: INFO @main_loop : Epoch 111 done 2019-10-09 06:37:15.718: INFO @main_loop : Training epoch 112 2019-10-09 06:40:25.743: INFO @log_variables: train loss mean: 0.269980 2019-10-09 06:40:25.743: INFO @log_variables: train age_loss mean: 4.510861 2019-10-09 06:40:25.743: INFO @log_variables: train gender_loss mean: 0.045055 2019-10-09 06:40:25.743: INFO @log_variables: train matching_loss nanmean: 0.340798 2019-10-09 06:40:25.743: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:40:25.743: INFO @log_variables: train age_mae mean: 4.987305 2019-10-09 06:40:25.743: INFO @log_variables: train gender_accuracy mean: 0.983927 2019-10-09 06:40:25.743: INFO @log_variables: train positive_distance nanmean: 0.752743 2019-10-09 06:40:25.743: INFO @log_variables: train negative_distance nanmean: 1.408467 2019-10-09 06:40:25.743: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:40:25.744: INFO @log_variables: valid loss mean: 0.466738 2019-10-09 06:40:25.744: INFO @log_variables: valid age_loss mean: 6.418109 2019-10-09 06:40:25.744: INFO @log_variables: valid gender_loss mean: 0.303009 2019-10-09 06:40:25.744: INFO @log_variables: valid matching_loss nanmean: 0.502069 2019-10-09 06:40:25.744: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:40:25.744: INFO @log_variables: valid age_mae mean: 6.900159 2019-10-09 06:40:25.744: INFO @log_variables: valid gender_accuracy mean: 0.915602 2019-10-09 06:40:25.744: INFO @log_variables: valid positive_distance nanmean: 0.791123 2019-10-09 06:40:25.744: INFO @log_variables: valid negative_distance nanmean: 1.376626 2019-10-09 06:40:25.744: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:40:25.744: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:40:27.911: INFO @metrics_hook: valid matching accuracy: 0.8883492234814415, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:40:28.339: INFO @decay_lr : LR updated to `0.0005704073` 2019-10-09 06:40:28.340: INFO @log_profile : T train: 179.801598 2019-10-09 06:40:28.340: INFO @log_profile : T valid: 8.492544 2019-10-09 06:40:28.340: INFO @log_profile : T read data: 1.019902 2019-10-09 06:40:28.340: INFO @log_profile : T hooks: 3.222215 2019-10-09 06:40:28.340: INFO @main_loop : Epoch 112 done 2019-10-09 06:40:28.340: INFO @main_loop : Training epoch 113 2019-10-09 06:43:38.572: INFO @log_variables: train loss mean: 0.270011 2019-10-09 06:43:38.572: INFO @log_variables: train age_loss mean: 4.498953 2019-10-09 06:43:38.572: INFO @log_variables: train gender_loss mean: 0.049501 2019-10-09 06:43:38.572: INFO @log_variables: train matching_loss nanmean: 0.337639 2019-10-09 06:43:38.573: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:43:38.573: INFO @log_variables: train age_mae mean: 4.975135 2019-10-09 06:43:38.573: INFO @log_variables: train gender_accuracy mean: 0.982064 2019-10-09 06:43:38.573: INFO @log_variables: train positive_distance nanmean: 0.752274 2019-10-09 06:43:38.573: INFO @log_variables: train negative_distance nanmean: 1.408485 2019-10-09 06:43:38.573: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:43:38.573: INFO @log_variables: valid loss mean: 0.457430 2019-10-09 06:43:38.573: INFO @log_variables: valid age_loss mean: 6.672353 2019-10-09 06:43:38.573: INFO @log_variables: valid gender_loss mean: 0.246374 2019-10-09 06:43:38.573: INFO @log_variables: valid matching_loss nanmean: 0.504425 2019-10-09 06:43:38.573: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:43:38.573: INFO @log_variables: valid age_mae mean: 7.157399 2019-10-09 06:43:38.573: INFO @log_variables: valid gender_accuracy mean: 0.920984 2019-10-09 06:43:38.573: INFO @log_variables: valid positive_distance nanmean: 0.782385 2019-10-09 06:43:38.573: INFO @log_variables: valid negative_distance nanmean: 1.373063 2019-10-09 06:43:38.573: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:43:38.573: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:43:40.823: INFO @metrics_hook: valid matching accuracy: 0.8905678479342808, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:43:41.279: INFO @decay_lr : LR updated to `0.00056755525` 2019-10-09 06:43:41.280: INFO @log_profile : T train: 179.610557 2019-10-09 06:43:41.280: INFO @log_profile : T valid: 8.501525 2019-10-09 06:43:41.280: INFO @log_profile : T read data: 1.441402 2019-10-09 06:43:41.280: INFO @log_profile : T hooks: 3.300000 2019-10-09 06:43:41.280: INFO @main_loop : Epoch 113 done 2019-10-09 06:43:41.280: INFO @main_loop : Training epoch 114 2019-10-09 06:46:51.490: INFO @log_variables: train loss mean: 0.268172 2019-10-09 06:46:51.490: INFO @log_variables: train age_loss mean: 4.488356 2019-10-09 06:46:51.490: INFO @log_variables: train gender_loss mean: 0.045903 2019-10-09 06:46:51.490: INFO @log_variables: train matching_loss nanmean: 0.336595 2019-10-09 06:46:51.490: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:46:51.491: INFO @log_variables: train age_mae mean: 4.964696 2019-10-09 06:46:51.491: INFO @log_variables: train gender_accuracy mean: 0.983736 2019-10-09 06:46:51.491: INFO @log_variables: train positive_distance nanmean: 0.751757 2019-10-09 06:46:51.491: INFO @log_variables: train negative_distance nanmean: 1.408289 2019-10-09 06:46:51.491: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:46:51.491: INFO @log_variables: valid loss mean: 0.458440 2019-10-09 06:46:51.491: INFO @log_variables: valid age_loss mean: 6.736649 2019-10-09 06:46:51.491: INFO @log_variables: valid gender_loss mean: 0.244817 2019-10-09 06:46:51.491: INFO @log_variables: valid matching_loss nanmean: 0.502684 2019-10-09 06:46:51.491: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:46:51.491: INFO @log_variables: valid age_mae mean: 7.220354 2019-10-09 06:46:51.491: INFO @log_variables: valid gender_accuracy mean: 0.925597 2019-10-09 06:46:51.491: INFO @log_variables: valid positive_distance nanmean: 0.788998 2019-10-09 06:46:51.491: INFO @log_variables: valid negative_distance nanmean: 1.377775 2019-10-09 06:46:51.491: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:46:51.491: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:46:53.670: INFO @metrics_hook: valid matching accuracy: 0.8884691491275409, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:46:54.103: INFO @decay_lr : LR updated to `0.00056471745` 2019-10-09 06:46:54.820: INFO @model : Quantizing and saving the model 2019-10-09 06:46:55.661: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 06:46:55.667: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 06:46:55.672: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 06:46:55.677: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 06:46:55.682: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 06:46:55.687: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 06:46:55.692: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 06:46:55.697: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 06:46:55.702: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 06:46:55.707: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 06:46:55.713: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 06:46:55.718: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 06:46:55.723: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 06:46:55.728: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 06:46:55.733: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 06:46:55.739: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 06:46:55.744: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 06:46:55.748: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 06:46:55.753: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 06:46:55.759: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 06:46:55.764: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 06:46:55.769: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 06:46:55.774: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 06:46:55.779: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 06:46:55.784: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 06:46:55.789: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 06:46:55.795: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 06:46:55.802: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 06:47:09.152: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 06:47:09.431: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 06:47:09.451: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 06:47:11.419: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 06:47:11.465: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 06:47:11.469: INFO @log_profile : T train: 180.009124 2019-10-09 06:47:11.470: INFO @log_profile : T valid: 8.500126 2019-10-09 06:47:11.470: INFO @log_profile : T read data: 1.018111 2019-10-09 06:47:11.471: INFO @log_profile : T hooks: 20.574707 2019-10-09 06:47:11.471: INFO @main_loop : Epoch 114 done 2019-10-09 06:47:11.471: INFO @main_loop : Training epoch 115 2019-10-09 06:50:20.196: INFO @log_variables: train loss mean: 0.270153 2019-10-09 06:50:20.197: INFO @log_variables: train age_loss mean: 4.481225 2019-10-09 06:50:20.197: INFO @log_variables: train gender_loss mean: 0.047670 2019-10-09 06:50:20.197: INFO @log_variables: train matching_loss nanmean: 0.341682 2019-10-09 06:50:20.197: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 06:50:20.197: INFO @log_variables: train age_mae mean: 4.957432 2019-10-09 06:50:20.197: INFO @log_variables: train gender_accuracy mean: 0.982883 2019-10-09 06:50:20.197: INFO @log_variables: train positive_distance nanmean: 0.754324 2019-10-09 06:50:20.197: INFO @log_variables: train negative_distance nanmean: 1.408144 2019-10-09 06:50:20.197: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:50:20.197: INFO @log_variables: valid loss mean: 0.458969 2019-10-09 06:50:20.197: INFO @log_variables: valid age_loss mean: 6.688827 2019-10-09 06:50:20.197: INFO @log_variables: valid gender_loss mean: 0.249259 2019-10-09 06:50:20.197: INFO @log_variables: valid matching_loss nanmean: 0.504661 2019-10-09 06:50:20.197: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:50:20.197: INFO @log_variables: valid age_mae mean: 7.172827 2019-10-09 06:50:20.197: INFO @log_variables: valid gender_accuracy mean: 0.924533 2019-10-09 06:50:20.197: INFO @log_variables: valid positive_distance nanmean: 0.789750 2019-10-09 06:50:20.198: INFO @log_variables: valid negative_distance nanmean: 1.371978 2019-10-09 06:50:20.198: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:50:20.198: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:50:21.966: INFO @metrics_hook: valid matching accuracy: 0.8879294837200935, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:50:22.390: INFO @decay_lr : LR updated to `0.00056189386` 2019-10-09 06:50:22.391: INFO @log_profile : T train: 178.271629 2019-10-09 06:50:22.391: INFO @log_profile : T valid: 8.317233 2019-10-09 06:50:22.392: INFO @log_profile : T read data: 1.456420 2019-10-09 06:50:22.392: INFO @log_profile : T hooks: 2.788178 2019-10-09 06:50:22.392: INFO @main_loop : Epoch 115 done 2019-10-09 06:50:22.392: INFO @main_loop : Training epoch 116 2019-10-09 06:53:32.917: INFO @log_variables: train loss mean: 0.269754 2019-10-09 06:53:32.917: INFO @log_variables: train age_loss mean: 4.480702 2019-10-09 06:53:32.917: INFO @log_variables: train gender_loss mean: 0.045300 2019-10-09 06:53:32.917: INFO @log_variables: train matching_loss nanmean: 0.342868 2019-10-09 06:53:32.917: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:53:32.917: INFO @log_variables: train age_mae mean: 4.957281 2019-10-09 06:53:32.917: INFO @log_variables: train gender_accuracy mean: 0.983717 2019-10-09 06:53:32.917: INFO @log_variables: train positive_distance nanmean: 0.754415 2019-10-09 06:53:32.917: INFO @log_variables: train negative_distance nanmean: 1.408658 2019-10-09 06:53:32.918: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:53:32.918: INFO @log_variables: valid loss mean: 0.450380 2019-10-09 06:53:32.918: INFO @log_variables: valid age_loss mean: 6.529051 2019-10-09 06:53:32.918: INFO @log_variables: valid gender_loss mean: 0.242739 2019-10-09 06:53:32.918: INFO @log_variables: valid matching_loss nanmean: 0.500535 2019-10-09 06:53:32.918: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:53:32.918: INFO @log_variables: valid age_mae mean: 7.012665 2019-10-09 06:53:32.918: INFO @log_variables: valid gender_accuracy mean: 0.924769 2019-10-09 06:53:32.918: INFO @log_variables: valid positive_distance nanmean: 0.783427 2019-10-09 06:53:32.918: INFO @log_variables: valid negative_distance nanmean: 1.374118 2019-10-09 06:53:32.918: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:53:32.918: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:53:35.215: INFO @metrics_hook: valid matching accuracy: 0.8883492234814415, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:53:35.640: INFO @decay_lr : LR updated to `0.0005590844` 2019-10-09 06:53:35.641: INFO @log_profile : T train: 179.957986 2019-10-09 06:53:35.641: INFO @log_profile : T valid: 8.483956 2019-10-09 06:53:35.641: INFO @log_profile : T read data: 1.407670 2019-10-09 06:53:35.641: INFO @log_profile : T hooks: 3.314651 2019-10-09 06:53:35.641: INFO @main_loop : Epoch 116 done 2019-10-09 06:53:35.642: INFO @main_loop : Training epoch 117 2019-10-09 06:56:45.509: INFO @log_variables: train loss mean: 0.266987 2019-10-09 06:56:45.509: INFO @log_variables: train age_loss mean: 4.449707 2019-10-09 06:56:45.509: INFO @log_variables: train gender_loss mean: 0.045583 2019-10-09 06:56:45.509: INFO @log_variables: train matching_loss nanmean: 0.337106 2019-10-09 06:56:45.510: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:56:45.510: INFO @log_variables: train age_mae mean: 4.926486 2019-10-09 06:56:45.510: INFO @log_variables: train gender_accuracy mean: 0.983683 2019-10-09 06:56:45.510: INFO @log_variables: train positive_distance nanmean: 0.752177 2019-10-09 06:56:45.510: INFO @log_variables: train negative_distance nanmean: 1.408600 2019-10-09 06:56:45.510: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:56:45.510: INFO @log_variables: valid loss mean: 0.457715 2019-10-09 06:56:45.510: INFO @log_variables: valid age_loss mean: 6.565248 2019-10-09 06:56:45.510: INFO @log_variables: valid gender_loss mean: 0.261949 2019-10-09 06:56:45.510: INFO @log_variables: valid matching_loss nanmean: 0.500442 2019-10-09 06:56:45.510: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:56:45.510: INFO @log_variables: valid age_mae mean: 7.048449 2019-10-09 06:56:45.510: INFO @log_variables: valid gender_accuracy mean: 0.920629 2019-10-09 06:56:45.510: INFO @log_variables: valid positive_distance nanmean: 0.782538 2019-10-09 06:56:45.510: INFO @log_variables: valid negative_distance nanmean: 1.373327 2019-10-09 06:56:45.510: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:56:45.510: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 06:56:47.630: INFO @metrics_hook: valid matching accuracy: 0.8887689632427894, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 06:56:48.056: INFO @decay_lr : LR updated to `0.000556289` 2019-10-09 06:56:48.057: INFO @log_profile : T train: 179.691844 2019-10-09 06:56:48.057: INFO @log_profile : T valid: 8.469833 2019-10-09 06:56:48.057: INFO @log_profile : T read data: 1.013461 2019-10-09 06:56:48.057: INFO @log_profile : T hooks: 3.153250 2019-10-09 06:56:48.057: INFO @main_loop : Epoch 117 done 2019-10-09 06:56:48.057: INFO @main_loop : Training epoch 118 2019-10-09 06:59:58.795: INFO @log_variables: train loss mean: 0.266548 2019-10-09 06:59:58.795: INFO @log_variables: train age_loss mean: 4.463166 2019-10-09 06:59:58.796: INFO @log_variables: train gender_loss mean: 0.044527 2019-10-09 06:59:58.796: INFO @log_variables: train matching_loss nanmean: 0.335454 2019-10-09 06:59:58.796: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 06:59:58.796: INFO @log_variables: train age_mae mean: 4.939392 2019-10-09 06:59:58.796: INFO @log_variables: train gender_accuracy mean: 0.984110 2019-10-09 06:59:58.796: INFO @log_variables: train positive_distance nanmean: 0.752112 2019-10-09 06:59:58.796: INFO @log_variables: train negative_distance nanmean: 1.408832 2019-10-09 06:59:58.796: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 06:59:58.796: INFO @log_variables: valid loss mean: 0.452556 2019-10-09 06:59:58.796: INFO @log_variables: valid age_loss mean: 6.410547 2019-10-09 06:59:58.796: INFO @log_variables: valid gender_loss mean: 0.259084 2019-10-09 06:59:58.796: INFO @log_variables: valid matching_loss nanmean: 0.502784 2019-10-09 06:59:58.796: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 06:59:58.796: INFO @log_variables: valid age_mae mean: 6.893594 2019-10-09 06:59:58.796: INFO @log_variables: valid gender_accuracy mean: 0.922581 2019-10-09 06:59:58.796: INFO @log_variables: valid positive_distance nanmean: 0.786567 2019-10-09 06:59:58.796: INFO @log_variables: valid negative_distance nanmean: 1.374905 2019-10-09 06:59:58.796: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 06:59:58.796: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:00:01.131: INFO @metrics_hook: valid matching accuracy: 0.8895484799424357, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:00:01.559: INFO @decay_lr : LR updated to `0.00055350753` 2019-10-09 07:00:01.561: INFO @log_profile : T train: 180.167041 2019-10-09 07:00:01.561: INFO @log_profile : T valid: 8.503075 2019-10-09 07:00:01.561: INFO @log_profile : T read data: 1.379310 2019-10-09 07:00:01.561: INFO @log_profile : T hooks: 3.367756 2019-10-09 07:00:01.561: INFO @main_loop : Epoch 118 done 2019-10-09 07:00:01.561: INFO @main_loop : Training epoch 119 2019-10-09 07:03:11.735: INFO @log_variables: train loss mean: 0.267342 2019-10-09 07:03:11.735: INFO @log_variables: train age_loss mean: 4.461453 2019-10-09 07:03:11.735: INFO @log_variables: train gender_loss mean: 0.046140 2019-10-09 07:03:11.735: INFO @log_variables: train matching_loss nanmean: 0.336476 2019-10-09 07:03:11.735: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:03:11.735: INFO @log_variables: train age_mae mean: 4.937827 2019-10-09 07:03:11.735: INFO @log_variables: train gender_accuracy mean: 0.983638 2019-10-09 07:03:11.735: INFO @log_variables: train positive_distance nanmean: 0.752590 2019-10-09 07:03:11.736: INFO @log_variables: train negative_distance nanmean: 1.408790 2019-10-09 07:03:11.736: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:03:11.736: INFO @log_variables: valid loss mean: 0.466447 2019-10-09 07:03:11.736: INFO @log_variables: valid age_loss mean: 6.810503 2019-10-09 07:03:11.736: INFO @log_variables: valid gender_loss mean: 0.263038 2019-10-09 07:03:11.736: INFO @log_variables: valid matching_loss nanmean: 0.501898 2019-10-09 07:03:11.736: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:03:11.736: INFO @log_variables: valid age_mae mean: 7.293943 2019-10-09 07:03:11.736: INFO @log_variables: valid gender_accuracy mean: 0.923527 2019-10-09 07:03:11.736: INFO @log_variables: valid positive_distance nanmean: 0.782266 2019-10-09 07:03:11.736: INFO @log_variables: valid negative_distance nanmean: 1.374156 2019-10-09 07:03:11.736: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:03:11.736: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:03:14.360: INFO @metrics_hook: valid matching accuracy: 0.8871499670204473, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:03:14.781: INFO @decay_lr : LR updated to `0.00055074` 2019-10-09 07:03:14.783: INFO @log_profile : T train: 179.580365 2019-10-09 07:03:14.783: INFO @log_profile : T valid: 8.491679 2019-10-09 07:03:14.783: INFO @log_profile : T read data: 1.451284 2019-10-09 07:03:14.783: INFO @log_profile : T hooks: 3.612289 2019-10-09 07:03:14.783: INFO @main_loop : Epoch 119 done 2019-10-09 07:03:14.783: INFO @main_loop : Training epoch 120 2019-10-09 07:06:25.439: INFO @log_variables: train loss mean: 0.265943 2019-10-09 07:06:25.439: INFO @log_variables: train age_loss mean: 4.437421 2019-10-09 07:06:25.439: INFO @log_variables: train gender_loss mean: 0.044191 2019-10-09 07:06:25.439: INFO @log_variables: train matching_loss nanmean: 0.336490 2019-10-09 07:06:25.439: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:06:25.439: INFO @log_variables: train age_mae mean: 4.913708 2019-10-09 07:06:25.439: INFO @log_variables: train gender_accuracy mean: 0.984053 2019-10-09 07:06:25.439: INFO @log_variables: train positive_distance nanmean: 0.752833 2019-10-09 07:06:25.439: INFO @log_variables: train negative_distance nanmean: 1.408842 2019-10-09 07:06:25.440: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:06:25.440: INFO @log_variables: valid loss mean: 0.448877 2019-10-09 07:06:25.440: INFO @log_variables: valid age_loss mean: 6.340049 2019-10-09 07:06:25.440: INFO @log_variables: valid gender_loss mean: 0.255580 2019-10-09 07:06:25.440: INFO @log_variables: valid matching_loss nanmean: 0.501933 2019-10-09 07:06:25.440: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:06:25.440: INFO @log_variables: valid age_mae mean: 6.822851 2019-10-09 07:06:25.440: INFO @log_variables: valid gender_accuracy mean: 0.924888 2019-10-09 07:06:25.440: INFO @log_variables: valid positive_distance nanmean: 0.784993 2019-10-09 07:06:25.440: INFO @log_variables: valid negative_distance nanmean: 1.373600 2019-10-09 07:06:25.440: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:06:25.440: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:06:27.662: INFO @metrics_hook: valid matching accuracy: 0.8883492234814415, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:06:28.112: INFO @decay_lr : LR updated to `0.0005479863` 2019-10-09 07:06:28.114: INFO @log_profile : T train: 180.427203 2019-10-09 07:06:28.114: INFO @log_profile : T valid: 8.533183 2019-10-09 07:06:28.114: INFO @log_profile : T read data: 0.987456 2019-10-09 07:06:28.114: INFO @log_profile : T hooks: 3.297862 2019-10-09 07:06:28.114: INFO @main_loop : Epoch 120 done 2019-10-09 07:06:28.114: INFO @main_loop : Training epoch 121 2019-10-09 07:09:38.517: INFO @log_variables: train loss mean: 0.264984 2019-10-09 07:09:38.517: INFO @log_variables: train age_loss mean: 4.405661 2019-10-09 07:09:38.517: INFO @log_variables: train gender_loss mean: 0.045979 2019-10-09 07:09:38.517: INFO @log_variables: train matching_loss nanmean: 0.334905 2019-10-09 07:09:38.517: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:09:38.517: INFO @log_variables: train age_mae mean: 4.881615 2019-10-09 07:09:38.517: INFO @log_variables: train gender_accuracy mean: 0.983366 2019-10-09 07:09:38.517: INFO @log_variables: train positive_distance nanmean: 0.751422 2019-10-09 07:09:38.517: INFO @log_variables: train negative_distance nanmean: 1.408544 2019-10-09 07:09:38.518: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:09:38.518: INFO @log_variables: valid loss mean: 0.450955 2019-10-09 07:09:38.518: INFO @log_variables: valid age_loss mean: 6.422774 2019-10-09 07:09:38.518: INFO @log_variables: valid gender_loss mean: 0.257367 2019-10-09 07:09:38.518: INFO @log_variables: valid matching_loss nanmean: 0.498314 2019-10-09 07:09:38.518: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:09:38.518: INFO @log_variables: valid age_mae mean: 6.905112 2019-10-09 07:09:38.518: INFO @log_variables: valid gender_accuracy mean: 0.923054 2019-10-09 07:09:38.518: INFO @log_variables: valid positive_distance nanmean: 0.786117 2019-10-09 07:09:38.518: INFO @log_variables: valid negative_distance nanmean: 1.375257 2019-10-09 07:09:38.518: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:09:38.518: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:09:40.753: INFO @metrics_hook: valid matching accuracy: 0.8897883312346345, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:09:41.179: INFO @decay_lr : LR updated to `0.00054524635` 2019-10-09 07:09:41.180: INFO @log_profile : T train: 179.818838 2019-10-09 07:09:41.180: INFO @log_profile : T valid: 8.515846 2019-10-09 07:09:41.180: INFO @log_profile : T read data: 1.429358 2019-10-09 07:09:41.180: INFO @log_profile : T hooks: 3.217762 2019-10-09 07:09:41.180: INFO @main_loop : Epoch 121 done 2019-10-09 07:09:41.180: INFO @main_loop : Training epoch 122 2019-10-09 07:12:51.439: INFO @log_variables: train loss mean: 0.265011 2019-10-09 07:12:51.439: INFO @log_variables: train age_loss mean: 4.431551 2019-10-09 07:12:51.439: INFO @log_variables: train gender_loss mean: 0.043860 2019-10-09 07:12:51.439: INFO @log_variables: train matching_loss nanmean: 0.334519 2019-10-09 07:12:51.439: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:12:51.439: INFO @log_variables: train age_mae mean: 4.907755 2019-10-09 07:12:51.439: INFO @log_variables: train gender_accuracy mean: 0.984604 2019-10-09 07:12:51.439: INFO @log_variables: train positive_distance nanmean: 0.749861 2019-10-09 07:12:51.439: INFO @log_variables: train negative_distance nanmean: 1.408514 2019-10-09 07:12:51.439: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:12:51.439: INFO @log_variables: valid loss mean: 0.451588 2019-10-09 07:12:51.439: INFO @log_variables: valid age_loss mean: 6.480137 2019-10-09 07:12:51.439: INFO @log_variables: valid gender_loss mean: 0.253480 2019-10-09 07:12:51.439: INFO @log_variables: valid matching_loss nanmean: 0.498428 2019-10-09 07:12:51.439: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:12:51.439: INFO @log_variables: valid age_mae mean: 6.962801 2019-10-09 07:12:51.440: INFO @log_variables: valid gender_accuracy mean: 0.924060 2019-10-09 07:12:51.440: INFO @log_variables: valid positive_distance nanmean: 0.787853 2019-10-09 07:12:51.440: INFO @log_variables: valid negative_distance nanmean: 1.372185 2019-10-09 07:12:51.440: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:12:51.440: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:12:53.679: INFO @metrics_hook: valid matching accuracy: 0.8902680338190322, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:12:54.085: INFO @decay_lr : LR updated to `0.00054252014` 2019-10-09 07:12:54.087: INFO @log_profile : T train: 180.021132 2019-10-09 07:12:54.087: INFO @log_profile : T valid: 8.580412 2019-10-09 07:12:54.087: INFO @log_profile : T read data: 0.979664 2019-10-09 07:12:54.087: INFO @log_profile : T hooks: 3.238929 2019-10-09 07:12:54.087: INFO @main_loop : Epoch 122 done 2019-10-09 07:12:54.087: INFO @main_loop : Training epoch 123 2019-10-09 07:16:04.810: INFO @log_variables: train loss mean: 0.263751 2019-10-09 07:16:04.810: INFO @log_variables: train age_loss mean: 4.414707 2019-10-09 07:16:04.810: INFO @log_variables: train gender_loss mean: 0.043214 2019-10-09 07:16:04.810: INFO @log_variables: train matching_loss nanmean: 0.332943 2019-10-09 07:16:04.811: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:16:04.811: INFO @log_variables: train age_mae mean: 4.890934 2019-10-09 07:16:04.811: INFO @log_variables: train gender_accuracy mean: 0.984314 2019-10-09 07:16:04.811: INFO @log_variables: train positive_distance nanmean: 0.750273 2019-10-09 07:16:04.811: INFO @log_variables: train negative_distance nanmean: 1.408735 2019-10-09 07:16:04.811: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:16:04.811: INFO @log_variables: valid loss mean: 0.454369 2019-10-09 07:16:04.811: INFO @log_variables: valid age_loss mean: 6.375163 2019-10-09 07:16:04.811: INFO @log_variables: valid gender_loss mean: 0.267837 2019-10-09 07:16:04.811: INFO @log_variables: valid matching_loss nanmean: 0.503191 2019-10-09 07:16:04.811: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:16:04.811: INFO @log_variables: valid age_mae mean: 6.858072 2019-10-09 07:16:04.811: INFO @log_variables: valid gender_accuracy mean: 0.920570 2019-10-09 07:16:04.811: INFO @log_variables: valid positive_distance nanmean: 0.780760 2019-10-09 07:16:04.812: INFO @log_variables: valid negative_distance nanmean: 1.372090 2019-10-09 07:16:04.812: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:16:04.812: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:16:07.152: INFO @metrics_hook: valid matching accuracy: 0.8866702644360497, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:16:07.588: INFO @decay_lr : LR updated to `0.00053980754` 2019-10-09 07:16:07.589: INFO @log_profile : T train: 180.060985 2019-10-09 07:16:07.589: INFO @log_profile : T valid: 8.604930 2019-10-09 07:16:07.590: INFO @log_profile : T read data: 1.394015 2019-10-09 07:16:07.590: INFO @log_profile : T hooks: 3.356333 2019-10-09 07:16:07.590: INFO @main_loop : Epoch 123 done 2019-10-09 07:16:07.590: INFO @main_loop : Training epoch 124 2019-10-09 07:19:18.413: INFO @log_variables: train loss mean: 0.264529 2019-10-09 07:19:18.413: INFO @log_variables: train age_loss mean: 4.407140 2019-10-09 07:19:18.413: INFO @log_variables: train gender_loss mean: 0.045331 2019-10-09 07:19:18.413: INFO @log_variables: train matching_loss nanmean: 0.333994 2019-10-09 07:19:18.413: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:19:18.413: INFO @log_variables: train age_mae mean: 4.883197 2019-10-09 07:19:18.413: INFO @log_variables: train gender_accuracy mean: 0.983295 2019-10-09 07:19:18.414: INFO @log_variables: train positive_distance nanmean: 0.751571 2019-10-09 07:19:18.414: INFO @log_variables: train negative_distance nanmean: 1.408504 2019-10-09 07:19:18.414: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:19:18.414: INFO @log_variables: valid loss mean: 0.450790 2019-10-09 07:19:18.414: INFO @log_variables: valid age_loss mean: 6.274205 2019-10-09 07:19:18.414: INFO @log_variables: valid gender_loss mean: 0.271345 2019-10-09 07:19:18.414: INFO @log_variables: valid matching_loss nanmean: 0.498683 2019-10-09 07:19:18.414: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:19:18.414: INFO @log_variables: valid age_mae mean: 6.756940 2019-10-09 07:19:18.414: INFO @log_variables: valid gender_accuracy mean: 0.920748 2019-10-09 07:19:18.414: INFO @log_variables: valid positive_distance nanmean: 0.786507 2019-10-09 07:19:18.414: INFO @log_variables: valid negative_distance nanmean: 1.375414 2019-10-09 07:19:18.414: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:19:18.414: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:19:20.816: INFO @metrics_hook: valid matching accuracy: 0.8896084427654855, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:19:21.231: INFO @decay_lr : LR updated to `0.0005371085` 2019-10-09 07:19:21.232: INFO @log_profile : T train: 180.147453 2019-10-09 07:19:21.232: INFO @log_profile : T valid: 8.599314 2019-10-09 07:19:21.232: INFO @log_profile : T read data: 1.404471 2019-10-09 07:19:21.232: INFO @log_profile : T hooks: 3.405143 2019-10-09 07:19:21.232: INFO @main_loop : Epoch 124 done 2019-10-09 07:19:21.232: INFO @main_loop : Training epoch 125 2019-10-09 07:22:31.308: INFO @log_variables: train loss mean: 0.264003 2019-10-09 07:22:31.309: INFO @log_variables: train age_loss mean: 4.405500 2019-10-09 07:22:31.309: INFO @log_variables: train gender_loss mean: 0.042132 2019-10-09 07:22:31.309: INFO @log_variables: train matching_loss nanmean: 0.335728 2019-10-09 07:22:31.309: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:22:31.309: INFO @log_variables: train age_mae mean: 4.881467 2019-10-09 07:22:31.309: INFO @log_variables: train gender_accuracy mean: 0.985293 2019-10-09 07:22:31.309: INFO @log_variables: train positive_distance nanmean: 0.752005 2019-10-09 07:22:31.309: INFO @log_variables: train negative_distance nanmean: 1.408668 2019-10-09 07:22:31.309: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:22:31.309: INFO @log_variables: valid loss mean: 0.455809 2019-10-09 07:22:31.309: INFO @log_variables: valid age_loss mean: 6.549482 2019-10-09 07:22:31.309: INFO @log_variables: valid gender_loss mean: 0.256642 2019-10-09 07:22:31.309: INFO @log_variables: valid matching_loss nanmean: 0.501419 2019-10-09 07:22:31.309: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:22:31.310: INFO @log_variables: valid age_mae mean: 7.033195 2019-10-09 07:22:31.310: INFO @log_variables: valid gender_accuracy mean: 0.921990 2019-10-09 07:22:31.310: INFO @log_variables: valid positive_distance nanmean: 0.784228 2019-10-09 07:22:31.310: INFO @log_variables: valid negative_distance nanmean: 1.374792 2019-10-09 07:22:31.310: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:22:31.310: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:22:33.588: INFO @metrics_hook: valid matching accuracy: 0.8874497811356958, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:22:34.033: INFO @decay_lr : LR updated to `0.000534423` 2019-10-09 07:22:34.035: INFO @log_profile : T train: 179.876770 2019-10-09 07:22:34.035: INFO @log_profile : T valid: 8.529200 2019-10-09 07:22:34.035: INFO @log_profile : T read data: 0.986297 2019-10-09 07:22:34.035: INFO @log_profile : T hooks: 3.323805 2019-10-09 07:22:34.035: INFO @main_loop : Epoch 125 done 2019-10-09 07:22:34.035: INFO @main_loop : Training epoch 126 2019-10-09 07:25:44.316: INFO @log_variables: train loss mean: 0.262583 2019-10-09 07:25:44.316: INFO @log_variables: train age_loss mean: 4.397328 2019-10-09 07:25:44.316: INFO @log_variables: train gender_loss mean: 0.042547 2019-10-09 07:25:44.317: INFO @log_variables: train matching_loss nanmean: 0.331726 2019-10-09 07:25:44.317: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:25:44.317: INFO @log_variables: train age_mae mean: 4.873219 2019-10-09 07:25:44.317: INFO @log_variables: train gender_accuracy mean: 0.985274 2019-10-09 07:25:44.317: INFO @log_variables: train positive_distance nanmean: 0.749850 2019-10-09 07:25:44.317: INFO @log_variables: train negative_distance nanmean: 1.408574 2019-10-09 07:25:44.317: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:25:44.317: INFO @log_variables: valid loss mean: 0.465462 2019-10-09 07:25:44.317: INFO @log_variables: valid age_loss mean: 6.476546 2019-10-09 07:25:44.317: INFO @log_variables: valid gender_loss mean: 0.289113 2019-10-09 07:25:44.317: INFO @log_variables: valid matching_loss nanmean: 0.506165 2019-10-09 07:25:44.317: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:25:44.317: INFO @log_variables: valid age_mae mean: 6.959574 2019-10-09 07:25:44.317: INFO @log_variables: valid gender_accuracy mean: 0.916371 2019-10-09 07:25:44.317: INFO @log_variables: valid positive_distance nanmean: 0.779491 2019-10-09 07:25:44.317: INFO @log_variables: valid negative_distance nanmean: 1.370034 2019-10-09 07:25:44.317: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:25:44.317: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:25:47.229: INFO @metrics_hook: valid matching accuracy: 0.88864903759669, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:25:47.663: INFO @decay_lr : LR updated to `0.00053175085` 2019-10-09 07:25:47.664: INFO @log_profile : T train: 179.718878 2019-10-09 07:25:47.664: INFO @log_profile : T valid: 8.552591 2019-10-09 07:25:47.664: INFO @log_profile : T read data: 1.371361 2019-10-09 07:25:47.664: INFO @log_profile : T hooks: 3.900658 2019-10-09 07:25:47.664: INFO @main_loop : Epoch 126 done 2019-10-09 07:25:47.664: INFO @main_loop : Training epoch 127 2019-10-09 07:28:57.209: INFO @log_variables: train loss mean: 0.264873 2019-10-09 07:28:57.209: INFO @log_variables: train age_loss mean: 4.415779 2019-10-09 07:28:57.209: INFO @log_variables: train gender_loss mean: 0.045342 2019-10-09 07:28:57.209: INFO @log_variables: train matching_loss nanmean: 0.334186 2019-10-09 07:28:57.209: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:28:57.209: INFO @log_variables: train age_mae mean: 4.891626 2019-10-09 07:28:57.209: INFO @log_variables: train gender_accuracy mean: 0.984242 2019-10-09 07:28:57.209: INFO @log_variables: train positive_distance nanmean: 0.750583 2019-10-09 07:28:57.209: INFO @log_variables: train negative_distance nanmean: 1.408473 2019-10-09 07:28:57.209: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:28:57.209: INFO @log_variables: valid loss mean: 0.467537 2019-10-09 07:28:57.209: INFO @log_variables: valid age_loss mean: 6.789622 2019-10-09 07:28:57.209: INFO @log_variables: valid gender_loss mean: 0.271717 2019-10-09 07:28:57.209: INFO @log_variables: valid matching_loss nanmean: 0.498686 2019-10-09 07:28:57.210: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:28:57.210: INFO @log_variables: valid age_mae mean: 7.274313 2019-10-09 07:28:57.210: INFO @log_variables: valid gender_accuracy mean: 0.921398 2019-10-09 07:28:57.210: INFO @log_variables: valid positive_distance nanmean: 0.786256 2019-10-09 07:28:57.210: INFO @log_variables: valid negative_distance nanmean: 1.372997 2019-10-09 07:28:57.210: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:28:57.210: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:28:59.575: INFO @metrics_hook: valid matching accuracy: 0.890327996642082, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:29:00.025: INFO @decay_lr : LR updated to `0.0005290921` 2019-10-09 07:29:00.026: INFO @log_profile : T train: 179.106140 2019-10-09 07:29:00.026: INFO @log_profile : T valid: 8.336897 2019-10-09 07:29:00.026: INFO @log_profile : T read data: 1.412660 2019-10-09 07:29:00.026: INFO @log_profile : T hooks: 3.418814 2019-10-09 07:29:00.026: INFO @main_loop : Epoch 127 done 2019-10-09 07:29:00.026: INFO @main_loop : Training epoch 128 2019-10-09 07:32:10.278: INFO @log_variables: train loss mean: 0.262950 2019-10-09 07:32:10.279: INFO @log_variables: train age_loss mean: 4.401523 2019-10-09 07:32:10.279: INFO @log_variables: train gender_loss mean: 0.041679 2019-10-09 07:32:10.279: INFO @log_variables: train matching_loss nanmean: 0.333312 2019-10-09 07:32:10.279: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:32:10.279: INFO @log_variables: train age_mae mean: 4.877668 2019-10-09 07:32:10.279: INFO @log_variables: train gender_accuracy mean: 0.984983 2019-10-09 07:32:10.279: INFO @log_variables: train positive_distance nanmean: 0.750811 2019-10-09 07:32:10.279: INFO @log_variables: train negative_distance nanmean: 1.408394 2019-10-09 07:32:10.279: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:32:10.279: INFO @log_variables: valid loss mean: 0.456911 2019-10-09 07:32:10.279: INFO @log_variables: valid age_loss mean: 6.393871 2019-10-09 07:32:10.279: INFO @log_variables: valid gender_loss mean: 0.269327 2019-10-09 07:32:10.279: INFO @log_variables: valid matching_loss nanmean: 0.507710 2019-10-09 07:32:10.279: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:32:10.279: INFO @log_variables: valid age_mae mean: 6.877409 2019-10-09 07:32:10.279: INFO @log_variables: valid gender_accuracy mean: 0.921457 2019-10-09 07:32:10.279: INFO @log_variables: valid positive_distance nanmean: 0.782526 2019-10-09 07:32:10.280: INFO @log_variables: valid negative_distance nanmean: 1.371450 2019-10-09 07:32:10.280: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:32:10.280: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:32:12.221: INFO @metrics_hook: valid matching accuracy: 0.8891287401810877, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:32:12.667: INFO @decay_lr : LR updated to `0.0005264466` 2019-10-09 07:32:12.668: INFO @log_profile : T train: 180.045179 2019-10-09 07:32:12.668: INFO @log_profile : T valid: 8.537148 2019-10-09 07:32:12.668: INFO @log_profile : T read data: 0.979154 2019-10-09 07:32:12.669: INFO @log_profile : T hooks: 2.995715 2019-10-09 07:32:12.669: INFO @main_loop : Epoch 128 done 2019-10-09 07:32:12.669: INFO @main_loop : Training epoch 129 2019-10-09 07:35:23.232: INFO @log_variables: train loss mean: 0.262494 2019-10-09 07:35:23.232: INFO @log_variables: train age_loss mean: 4.388506 2019-10-09 07:35:23.232: INFO @log_variables: train gender_loss mean: 0.042518 2019-10-09 07:35:23.232: INFO @log_variables: train matching_loss nanmean: 0.332364 2019-10-09 07:35:23.232: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:35:23.232: INFO @log_variables: train age_mae mean: 4.864702 2019-10-09 07:35:23.232: INFO @log_variables: train gender_accuracy mean: 0.984805 2019-10-09 07:35:23.232: INFO @log_variables: train positive_distance nanmean: 0.749659 2019-10-09 07:35:23.232: INFO @log_variables: train negative_distance nanmean: 1.408405 2019-10-09 07:35:23.232: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:35:23.232: INFO @log_variables: valid loss mean: 0.450938 2019-10-09 07:35:23.232: INFO @log_variables: valid age_loss mean: 6.491864 2019-10-09 07:35:23.232: INFO @log_variables: valid gender_loss mean: 0.248673 2019-10-09 07:35:23.232: INFO @log_variables: valid matching_loss nanmean: 0.500049 2019-10-09 07:35:23.232: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:35:23.233: INFO @log_variables: valid age_mae mean: 6.974681 2019-10-09 07:35:23.233: INFO @log_variables: valid gender_accuracy mean: 0.924769 2019-10-09 07:35:23.233: INFO @log_variables: valid positive_distance nanmean: 0.782109 2019-10-09 07:35:23.233: INFO @log_variables: valid negative_distance nanmean: 1.372488 2019-10-09 07:35:23.233: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:35:23.233: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:35:25.628: INFO @metrics_hook: valid matching accuracy: 0.8875097439587456, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:35:26.044: INFO @decay_lr : LR updated to `0.0005238144` 2019-10-09 07:35:26.046: INFO @log_profile : T train: 179.971427 2019-10-09 07:35:26.046: INFO @log_profile : T valid: 8.556518 2019-10-09 07:35:26.046: INFO @log_profile : T read data: 1.378086 2019-10-09 07:35:26.046: INFO @log_profile : T hooks: 3.386486 2019-10-09 07:35:26.047: INFO @main_loop : Epoch 129 done 2019-10-09 07:35:26.047: INFO @main_loop : Training epoch 130 2019-10-09 07:38:36.028: INFO @log_variables: train loss mean: 0.261895 2019-10-09 07:38:36.028: INFO @log_variables: train age_loss mean: 4.376630 2019-10-09 07:38:36.028: INFO @log_variables: train gender_loss mean: 0.042368 2019-10-09 07:38:36.028: INFO @log_variables: train matching_loss nanmean: 0.331844 2019-10-09 07:38:36.028: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:38:36.028: INFO @log_variables: train age_mae mean: 4.852546 2019-10-09 07:38:36.028: INFO @log_variables: train gender_accuracy mean: 0.984927 2019-10-09 07:38:36.029: INFO @log_variables: train positive_distance nanmean: 0.750401 2019-10-09 07:38:36.029: INFO @log_variables: train negative_distance nanmean: 1.408392 2019-10-09 07:38:36.029: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:38:36.029: INFO @log_variables: valid loss mean: 0.458011 2019-10-09 07:38:36.029: INFO @log_variables: valid age_loss mean: 6.476235 2019-10-09 07:38:36.029: INFO @log_variables: valid gender_loss mean: 0.273346 2019-10-09 07:38:36.029: INFO @log_variables: valid matching_loss nanmean: 0.498863 2019-10-09 07:38:36.029: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:38:36.029: INFO @log_variables: valid age_mae mean: 6.959454 2019-10-09 07:38:36.029: INFO @log_variables: valid gender_accuracy mean: 0.927135 2019-10-09 07:38:36.029: INFO @log_variables: valid positive_distance nanmean: 0.789533 2019-10-09 07:38:36.029: INFO @log_variables: valid negative_distance nanmean: 1.376275 2019-10-09 07:38:36.029: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:38:36.029: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:38:38.634: INFO @metrics_hook: valid matching accuracy: 0.889068777358038, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:38:39.080: INFO @decay_lr : LR updated to `0.00052119535` 2019-10-09 07:38:39.801: INFO @model : Quantizing and saving the model 2019-10-09 07:38:41.039: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 07:38:41.045: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 07:38:41.051: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 07:38:41.057: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 07:38:41.063: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 07:38:41.068: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 07:38:41.074: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 07:38:41.080: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 07:38:41.086: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 07:38:41.092: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 07:38:41.098: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 07:38:41.104: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 07:38:41.111: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 07:38:41.118: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 07:38:41.124: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 07:38:41.130: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 07:38:41.136: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 07:38:41.141: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 07:38:41.147: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 07:38:41.153: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 07:38:41.158: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 07:38:41.164: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 07:38:41.170: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 07:38:41.175: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 07:38:41.181: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 07:38:41.186: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 07:38:41.192: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 07:38:41.200: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 07:38:56.285: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 07:38:56.587: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 07:38:56.606: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 07:38:58.623: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 07:38:58.675: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 07:38:58.679: INFO @log_profile : T train: 179.797534 2019-10-09 07:38:58.682: INFO @log_profile : T valid: 8.565779 2019-10-09 07:38:58.683: INFO @log_profile : T read data: 0.980954 2019-10-09 07:38:58.683: INFO @log_profile : T hooks: 23.200323 2019-10-09 07:38:58.683: INFO @main_loop : Epoch 130 done 2019-10-09 07:38:58.683: INFO @main_loop : Training epoch 131 2019-10-09 07:42:09.111: INFO @log_variables: train loss mean: 0.260857 2019-10-09 07:42:09.111: INFO @log_variables: train age_loss mean: 4.391700 2019-10-09 07:42:09.111: INFO @log_variables: train gender_loss mean: 0.040838 2019-10-09 07:42:09.111: INFO @log_variables: train matching_loss nanmean: 0.328648 2019-10-09 07:42:09.111: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:42:09.111: INFO @log_variables: train age_mae mean: 4.867666 2019-10-09 07:42:09.111: INFO @log_variables: train gender_accuracy mean: 0.985183 2019-10-09 07:42:09.111: INFO @log_variables: train positive_distance nanmean: 0.747765 2019-10-09 07:42:09.111: INFO @log_variables: train negative_distance nanmean: 1.408515 2019-10-09 07:42:09.111: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:42:09.111: INFO @log_variables: valid loss mean: 0.453765 2019-10-09 07:42:09.111: INFO @log_variables: valid age_loss mean: 6.246035 2019-10-09 07:42:09.111: INFO @log_variables: valid gender_loss mean: 0.282157 2019-10-09 07:42:09.112: INFO @log_variables: valid matching_loss nanmean: 0.499912 2019-10-09 07:42:09.112: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:42:09.112: INFO @log_variables: valid age_mae mean: 6.727767 2019-10-09 07:42:09.112: INFO @log_variables: valid gender_accuracy mean: 0.917790 2019-10-09 07:42:09.112: INFO @log_variables: valid positive_distance nanmean: 0.781820 2019-10-09 07:42:09.112: INFO @log_variables: valid negative_distance nanmean: 1.372784 2019-10-09 07:42:09.112: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:42:09.112: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:42:11.604: INFO @metrics_hook: valid matching accuracy: 0.8891287401810877, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:42:12.081: INFO @decay_lr : LR updated to `0.0005185894` 2019-10-09 07:42:12.082: INFO @log_profile : T train: 179.645826 2019-10-09 07:42:12.082: INFO @log_profile : T valid: 8.623123 2019-10-09 07:42:12.082: INFO @log_profile : T read data: 1.455525 2019-10-09 07:42:12.082: INFO @log_profile : T hooks: 3.588153 2019-10-09 07:42:12.082: INFO @main_loop : Epoch 131 done 2019-10-09 07:42:12.082: INFO @main_loop : Training epoch 132 2019-10-09 07:45:22.383: INFO @log_variables: train loss mean: 0.261442 2019-10-09 07:45:22.383: INFO @log_variables: train age_loss mean: 4.364794 2019-10-09 07:45:22.383: INFO @log_variables: train gender_loss mean: 0.042204 2019-10-09 07:45:22.383: INFO @log_variables: train matching_loss nanmean: 0.331786 2019-10-09 07:45:22.383: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:45:22.383: INFO @log_variables: train age_mae mean: 4.840244 2019-10-09 07:45:22.383: INFO @log_variables: train gender_accuracy mean: 0.985129 2019-10-09 07:45:22.383: INFO @log_variables: train positive_distance nanmean: 0.748886 2019-10-09 07:45:22.383: INFO @log_variables: train negative_distance nanmean: 1.408601 2019-10-09 07:45:22.383: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:45:22.383: INFO @log_variables: valid loss mean: 0.463724 2019-10-09 07:45:22.383: INFO @log_variables: valid age_loss mean: 6.517515 2019-10-09 07:45:22.383: INFO @log_variables: valid gender_loss mean: 0.289801 2019-10-09 07:45:22.383: INFO @log_variables: valid matching_loss nanmean: 0.495991 2019-10-09 07:45:22.384: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:45:22.384: INFO @log_variables: valid age_mae mean: 7.001208 2019-10-09 07:45:22.384: INFO @log_variables: valid gender_accuracy mean: 0.920629 2019-10-09 07:45:22.384: INFO @log_variables: valid positive_distance nanmean: 0.792177 2019-10-09 07:45:22.384: INFO @log_variables: valid negative_distance nanmean: 1.378402 2019-10-09 07:45:22.384: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:45:22.384: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:45:24.630: INFO @metrics_hook: valid matching accuracy: 0.8885890747736404, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:45:25.163: INFO @decay_lr : LR updated to `0.0005159965` 2019-10-09 07:45:25.165: INFO @log_profile : T train: 179.633676 2019-10-09 07:45:25.165: INFO @log_profile : T valid: 8.535289 2019-10-09 07:45:25.165: INFO @log_profile : T read data: 1.454472 2019-10-09 07:45:25.165: INFO @log_profile : T hooks: 3.374842 2019-10-09 07:45:25.165: INFO @main_loop : Epoch 132 done 2019-10-09 07:45:25.165: INFO @main_loop : Training epoch 133 2019-10-09 07:48:35.138: INFO @log_variables: train loss mean: 0.262176 2019-10-09 07:48:35.138: INFO @log_variables: train age_loss mean: 4.364147 2019-10-09 07:48:35.138: INFO @log_variables: train gender_loss mean: 0.045716 2019-10-09 07:48:35.138: INFO @log_variables: train matching_loss nanmean: 0.330615 2019-10-09 07:48:35.139: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 07:48:35.139: INFO @log_variables: train age_mae mean: 4.840308 2019-10-09 07:48:35.139: INFO @log_variables: train gender_accuracy mean: 0.984098 2019-10-09 07:48:35.139: INFO @log_variables: train positive_distance nanmean: 0.749598 2019-10-09 07:48:35.139: INFO @log_variables: train negative_distance nanmean: 1.408679 2019-10-09 07:48:35.139: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:48:35.139: INFO @log_variables: valid loss mean: 0.451365 2019-10-09 07:48:35.139: INFO @log_variables: valid age_loss mean: 6.365120 2019-10-09 07:48:35.139: INFO @log_variables: valid gender_loss mean: 0.270671 2019-10-09 07:48:35.139: INFO @log_variables: valid matching_loss nanmean: 0.492047 2019-10-09 07:48:35.139: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:48:35.139: INFO @log_variables: valid age_mae mean: 6.848786 2019-10-09 07:48:35.139: INFO @log_variables: valid gender_accuracy mean: 0.921812 2019-10-09 07:48:35.139: INFO @log_variables: valid positive_distance nanmean: 0.786041 2019-10-09 07:48:35.139: INFO @log_variables: valid negative_distance nanmean: 1.376121 2019-10-09 07:48:35.139: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:48:35.139: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:48:37.266: INFO @metrics_hook: valid matching accuracy: 0.8899682197037837, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:48:37.702: INFO @decay_lr : LR updated to `0.0005134165` 2019-10-09 07:48:37.704: INFO @log_profile : T train: 179.784689 2019-10-09 07:48:37.704: INFO @log_profile : T valid: 8.511581 2019-10-09 07:48:37.704: INFO @log_profile : T read data: 1.007281 2019-10-09 07:48:37.704: INFO @log_profile : T hooks: 3.150082 2019-10-09 07:48:37.704: INFO @main_loop : Epoch 133 done 2019-10-09 07:48:37.704: INFO @main_loop : Training epoch 134 2019-10-09 07:51:48.156: INFO @log_variables: train loss mean: 0.260767 2019-10-09 07:51:48.157: INFO @log_variables: train age_loss mean: 4.352922 2019-10-09 07:51:48.157: INFO @log_variables: train gender_loss mean: 0.040975 2019-10-09 07:51:48.157: INFO @log_variables: train matching_loss nanmean: 0.332110 2019-10-09 07:51:48.157: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:51:48.157: INFO @log_variables: train age_mae mean: 4.828688 2019-10-09 07:51:48.157: INFO @log_variables: train gender_accuracy mean: 0.985503 2019-10-09 07:51:48.157: INFO @log_variables: train positive_distance nanmean: 0.749666 2019-10-09 07:51:48.157: INFO @log_variables: train negative_distance nanmean: 1.408522 2019-10-09 07:51:48.157: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:51:48.157: INFO @log_variables: valid loss mean: 0.464359 2019-10-09 07:51:48.157: INFO @log_variables: valid age_loss mean: 6.630352 2019-10-09 07:51:48.157: INFO @log_variables: valid gender_loss mean: 0.273165 2019-10-09 07:51:48.157: INFO @log_variables: valid matching_loss nanmean: 0.503314 2019-10-09 07:51:48.157: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:51:48.157: INFO @log_variables: valid age_mae mean: 7.114360 2019-10-09 07:51:48.158: INFO @log_variables: valid gender_accuracy mean: 0.923705 2019-10-09 07:51:48.158: INFO @log_variables: valid positive_distance nanmean: 0.784751 2019-10-09 07:51:48.158: INFO @log_variables: valid negative_distance nanmean: 1.371961 2019-10-09 07:51:48.158: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:51:48.158: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:51:50.364: INFO @metrics_hook: valid matching accuracy: 0.8912874018108773, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:51:50.812: INFO @decay_lr : LR updated to `0.0005108494` 2019-10-09 07:51:50.813: INFO @log_profile : T train: 179.869337 2019-10-09 07:51:50.813: INFO @log_profile : T valid: 8.506592 2019-10-09 07:51:50.813: INFO @log_profile : T read data: 1.438894 2019-10-09 07:51:50.814: INFO @log_profile : T hooks: 3.210238 2019-10-09 07:51:50.814: INFO @main_loop : Epoch 134 done 2019-10-09 07:51:50.814: INFO @main_loop : Training epoch 135 2019-10-09 07:55:01.160: INFO @log_variables: train loss mean: 0.258269 2019-10-09 07:55:01.160: INFO @log_variables: train age_loss mean: 4.289817 2019-10-09 07:55:01.160: INFO @log_variables: train gender_loss mean: 0.042629 2019-10-09 07:55:01.160: INFO @log_variables: train matching_loss nanmean: 0.329024 2019-10-09 07:55:01.160: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:55:01.160: INFO @log_variables: train age_mae mean: 4.765252 2019-10-09 07:55:01.160: INFO @log_variables: train gender_accuracy mean: 0.984895 2019-10-09 07:55:01.160: INFO @log_variables: train positive_distance nanmean: 0.748882 2019-10-09 07:55:01.160: INFO @log_variables: train negative_distance nanmean: 1.408651 2019-10-09 07:55:01.160: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:55:01.160: INFO @log_variables: valid loss mean: 0.455936 2019-10-09 07:55:01.160: INFO @log_variables: valid age_loss mean: 6.550395 2019-10-09 07:55:01.160: INFO @log_variables: valid gender_loss mean: 0.259309 2019-10-09 07:55:01.160: INFO @log_variables: valid matching_loss nanmean: 0.499053 2019-10-09 07:55:01.160: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:55:01.161: INFO @log_variables: valid age_mae mean: 7.034385 2019-10-09 07:55:01.161: INFO @log_variables: valid gender_accuracy mean: 0.926721 2019-10-09 07:55:01.161: INFO @log_variables: valid positive_distance nanmean: 0.786136 2019-10-09 07:55:01.161: INFO @log_variables: valid negative_distance nanmean: 1.375367 2019-10-09 07:55:01.161: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:55:01.161: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:55:03.562: INFO @metrics_hook: valid matching accuracy: 0.8882292978353421, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:55:03.995: INFO @decay_lr : LR updated to `0.00050829514` 2019-10-09 07:55:03.996: INFO @log_profile : T train: 179.751161 2019-10-09 07:55:03.996: INFO @log_profile : T valid: 8.488246 2019-10-09 07:55:03.996: INFO @log_profile : T read data: 1.441959 2019-10-09 07:55:03.996: INFO @log_profile : T hooks: 3.416589 2019-10-09 07:55:03.996: INFO @main_loop : Epoch 135 done 2019-10-09 07:55:03.996: INFO @main_loop : Training epoch 136 2019-10-09 07:58:14.187: INFO @log_variables: train loss mean: 0.260105 2019-10-09 07:58:14.188: INFO @log_variables: train age_loss mean: 4.328393 2019-10-09 07:58:14.188: INFO @log_variables: train gender_loss mean: 0.042884 2019-10-09 07:58:14.188: INFO @log_variables: train matching_loss nanmean: 0.330603 2019-10-09 07:58:14.188: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 07:58:14.188: INFO @log_variables: train age_mae mean: 4.803771 2019-10-09 07:58:14.188: INFO @log_variables: train gender_accuracy mean: 0.984578 2019-10-09 07:58:14.188: INFO @log_variables: train positive_distance nanmean: 0.750525 2019-10-09 07:58:14.188: INFO @log_variables: train negative_distance nanmean: 1.408250 2019-10-09 07:58:14.188: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 07:58:14.188: INFO @log_variables: valid loss mean: 0.461024 2019-10-09 07:58:14.188: INFO @log_variables: valid age_loss mean: 6.705195 2019-10-09 07:58:14.188: INFO @log_variables: valid gender_loss mean: 0.249994 2019-10-09 07:58:14.188: INFO @log_variables: valid matching_loss nanmean: 0.508661 2019-10-09 07:58:14.188: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 07:58:14.188: INFO @log_variables: valid age_mae mean: 7.189770 2019-10-09 07:58:14.188: INFO @log_variables: valid gender_accuracy mean: 0.925361 2019-10-09 07:58:14.189: INFO @log_variables: valid positive_distance nanmean: 0.781009 2019-10-09 07:58:14.189: INFO @log_variables: valid negative_distance nanmean: 1.368191 2019-10-09 07:58:14.189: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 07:58:14.189: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 07:58:16.673: INFO @metrics_hook: valid matching accuracy: 0.8868501529051988, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 07:58:17.130: INFO @decay_lr : LR updated to `0.0005057537` 2019-10-09 07:58:17.132: INFO @log_profile : T train: 179.958340 2019-10-09 07:58:17.132: INFO @log_profile : T valid: 8.506829 2019-10-09 07:58:17.132: INFO @log_profile : T read data: 1.048854 2019-10-09 07:58:17.132: INFO @log_profile : T hooks: 3.535413 2019-10-09 07:58:17.132: INFO @main_loop : Epoch 136 done 2019-10-09 07:58:17.132: INFO @main_loop : Training epoch 137 2019-10-09 08:01:27.645: INFO @log_variables: train loss mean: 0.260226 2019-10-09 08:01:27.645: INFO @log_variables: train age_loss mean: 4.332797 2019-10-09 08:01:27.645: INFO @log_variables: train gender_loss mean: 0.041643 2019-10-09 08:01:27.645: INFO @log_variables: train matching_loss nanmean: 0.331779 2019-10-09 08:01:27.645: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:01:27.645: INFO @log_variables: train age_mae mean: 4.808521 2019-10-09 08:01:27.645: INFO @log_variables: train gender_accuracy mean: 0.984867 2019-10-09 08:01:27.645: INFO @log_variables: train positive_distance nanmean: 0.751578 2019-10-09 08:01:27.645: INFO @log_variables: train negative_distance nanmean: 1.408510 2019-10-09 08:01:27.645: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:01:27.645: INFO @log_variables: valid loss mean: 0.449469 2019-10-09 08:01:27.645: INFO @log_variables: valid age_loss mean: 6.362213 2019-10-09 08:01:27.646: INFO @log_variables: valid gender_loss mean: 0.253217 2019-10-09 08:01:27.646: INFO @log_variables: valid matching_loss nanmean: 0.503916 2019-10-09 08:01:27.646: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:01:27.646: INFO @log_variables: valid age_mae mean: 6.846242 2019-10-09 08:01:27.646: INFO @log_variables: valid gender_accuracy mean: 0.926011 2019-10-09 08:01:27.646: INFO @log_variables: valid positive_distance nanmean: 0.785734 2019-10-09 08:01:27.646: INFO @log_variables: valid negative_distance nanmean: 1.376679 2019-10-09 08:01:27.646: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:01:27.646: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:01:29.461: INFO @metrics_hook: valid matching accuracy: 0.8891287401810877, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:01:29.891: INFO @decay_lr : LR updated to `0.0005032249` 2019-10-09 08:01:29.892: INFO @log_profile : T train: 179.853506 2019-10-09 08:01:29.892: INFO @log_profile : T valid: 8.516362 2019-10-09 08:01:29.892: INFO @log_profile : T read data: 1.464440 2019-10-09 08:01:29.892: INFO @log_profile : T hooks: 2.840819 2019-10-09 08:01:29.892: INFO @main_loop : Epoch 137 done 2019-10-09 08:01:29.892: INFO @main_loop : Training epoch 138 2019-10-09 08:04:40.108: INFO @log_variables: train loss mean: 0.259258 2019-10-09 08:04:40.109: INFO @log_variables: train age_loss mean: 4.333691 2019-10-09 08:04:40.109: INFO @log_variables: train gender_loss mean: 0.041983 2019-10-09 08:04:40.109: INFO @log_variables: train matching_loss nanmean: 0.328346 2019-10-09 08:04:40.109: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:04:40.109: INFO @log_variables: train age_mae mean: 4.809326 2019-10-09 08:04:40.109: INFO @log_variables: train gender_accuracy mean: 0.985338 2019-10-09 08:04:40.109: INFO @log_variables: train positive_distance nanmean: 0.747610 2019-10-09 08:04:40.109: INFO @log_variables: train negative_distance nanmean: 1.408435 2019-10-09 08:04:40.109: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:04:40.109: INFO @log_variables: valid loss mean: 0.446194 2019-10-09 08:04:40.109: INFO @log_variables: valid age_loss mean: 6.430276 2019-10-09 08:04:40.109: INFO @log_variables: valid gender_loss mean: 0.238086 2019-10-09 08:04:40.109: INFO @log_variables: valid matching_loss nanmean: 0.502089 2019-10-09 08:04:40.109: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:04:40.109: INFO @log_variables: valid age_mae mean: 6.913059 2019-10-09 08:04:40.110: INFO @log_variables: valid gender_accuracy mean: 0.928732 2019-10-09 08:04:40.110: INFO @log_variables: valid positive_distance nanmean: 0.784596 2019-10-09 08:04:40.110: INFO @log_variables: valid negative_distance nanmean: 1.373432 2019-10-09 08:04:40.110: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:04:40.110: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:04:42.526: INFO @metrics_hook: valid matching accuracy: 0.8878095580739941, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:04:42.963: INFO @decay_lr : LR updated to `0.00050070876` 2019-10-09 08:04:43.698: INFO @model : Quantizing and saving the model 2019-10-09 08:04:44.533: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 08:04:44.539: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 08:04:44.544: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 08:04:44.549: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 08:04:44.554: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 08:04:44.559: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 08:04:44.565: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 08:04:44.570: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 08:04:44.575: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 08:04:44.580: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 08:04:44.586: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 08:04:44.591: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 08:04:44.596: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 08:04:44.601: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 08:04:44.606: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 08:04:44.611: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 08:04:44.616: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 08:04:44.622: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 08:04:44.627: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 08:04:44.633: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 08:04:44.638: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 08:04:44.644: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 08:04:44.649: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 08:04:44.654: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 08:04:44.659: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 08:04:44.664: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 08:04:44.669: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 08:04:44.676: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 08:04:58.294: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 08:04:58.575: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 08:04:58.594: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 08:05:00.630: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 08:05:00.682: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 08:05:00.687: INFO @log_profile : T train: 179.990448 2019-10-09 08:05:00.687: INFO @log_profile : T valid: 8.529425 2019-10-09 08:05:00.687: INFO @log_profile : T read data: 1.041903 2019-10-09 08:05:00.687: INFO @log_profile : T hooks: 21.146295 2019-10-09 08:05:00.688: INFO @main_loop : Epoch 138 done 2019-10-09 08:05:00.688: INFO @main_loop : Training epoch 139 2019-10-09 08:08:10.840: INFO @log_variables: train loss mean: 0.258427 2019-10-09 08:08:10.841: INFO @log_variables: train age_loss mean: 4.319273 2019-10-09 08:08:10.841: INFO @log_variables: train gender_loss mean: 0.040374 2019-10-09 08:08:10.841: INFO @log_variables: train matching_loss nanmean: 0.328822 2019-10-09 08:08:10.841: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:08:10.841: INFO @log_variables: train age_mae mean: 4.794697 2019-10-09 08:08:10.841: INFO @log_variables: train gender_accuracy mean: 0.985753 2019-10-09 08:08:10.841: INFO @log_variables: train positive_distance nanmean: 0.748083 2019-10-09 08:08:10.841: INFO @log_variables: train negative_distance nanmean: 1.408179 2019-10-09 08:08:10.841: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:08:10.841: INFO @log_variables: valid loss mean: 0.449372 2019-10-09 08:08:10.841: INFO @log_variables: valid age_loss mean: 6.435616 2019-10-09 08:08:10.841: INFO @log_variables: valid gender_loss mean: 0.251917 2019-10-09 08:08:10.841: INFO @log_variables: valid matching_loss nanmean: 0.497576 2019-10-09 08:08:10.841: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:08:10.841: INFO @log_variables: valid age_mae mean: 6.918051 2019-10-09 08:08:10.841: INFO @log_variables: valid gender_accuracy mean: 0.926071 2019-10-09 08:08:10.841: INFO @log_variables: valid positive_distance nanmean: 0.788674 2019-10-09 08:08:10.842: INFO @log_variables: valid negative_distance nanmean: 1.376857 2019-10-09 08:08:10.842: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:08:10.842: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:08:13.141: INFO @metrics_hook: valid matching accuracy: 0.8901481081729328, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:08:13.577: INFO @decay_lr : LR updated to `0.0004982052` 2019-10-09 08:08:13.578: INFO @log_profile : T train: 179.549015 2019-10-09 08:08:13.578: INFO @log_profile : T valid: 8.517466 2019-10-09 08:08:13.578: INFO @log_profile : T read data: 1.401334 2019-10-09 08:08:13.578: INFO @log_profile : T hooks: 3.334463 2019-10-09 08:08:13.578: INFO @main_loop : Epoch 139 done 2019-10-09 08:08:13.578: INFO @main_loop : Training epoch 140 2019-10-09 08:11:24.119: INFO @log_variables: train loss mean: 0.258539 2019-10-09 08:11:24.119: INFO @log_variables: train age_loss mean: 4.326858 2019-10-09 08:11:24.119: INFO @log_variables: train gender_loss mean: 0.042116 2019-10-09 08:11:24.120: INFO @log_variables: train matching_loss nanmean: 0.326671 2019-10-09 08:11:24.120: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:11:24.120: INFO @log_variables: train age_mae mean: 4.802659 2019-10-09 08:11:24.120: INFO @log_variables: train gender_accuracy mean: 0.984974 2019-10-09 08:11:24.120: INFO @log_variables: train positive_distance nanmean: 0.745539 2019-10-09 08:11:24.120: INFO @log_variables: train negative_distance nanmean: 1.407814 2019-10-09 08:11:24.120: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:11:24.120: INFO @log_variables: valid loss mean: 0.454881 2019-10-09 08:11:24.120: INFO @log_variables: valid age_loss mean: 6.601772 2019-10-09 08:11:24.120: INFO @log_variables: valid gender_loss mean: 0.249919 2019-10-09 08:11:24.120: INFO @log_variables: valid matching_loss nanmean: 0.500034 2019-10-09 08:11:24.120: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:11:24.120: INFO @log_variables: valid age_mae mean: 7.085392 2019-10-09 08:11:24.120: INFO @log_variables: valid gender_accuracy mean: 0.927490 2019-10-09 08:11:24.120: INFO @log_variables: valid positive_distance nanmean: 0.793339 2019-10-09 08:11:24.120: INFO @log_variables: valid negative_distance nanmean: 1.376279 2019-10-09 08:11:24.120: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:11:24.121: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:11:26.340: INFO @metrics_hook: valid matching accuracy: 0.8894285542963363, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:11:26.767: INFO @decay_lr : LR updated to `0.00049571414` 2019-10-09 08:11:26.768: INFO @log_profile : T train: 179.924402 2019-10-09 08:11:26.768: INFO @log_profile : T valid: 8.516550 2019-10-09 08:11:26.768: INFO @log_profile : T read data: 1.430733 2019-10-09 08:11:26.768: INFO @log_profile : T hooks: 3.233692 2019-10-09 08:11:26.769: INFO @main_loop : Epoch 140 done 2019-10-09 08:11:26.769: INFO @main_loop : Training epoch 141 2019-10-09 08:14:36.661: INFO @log_variables: train loss mean: 0.256541 2019-10-09 08:14:36.661: INFO @log_variables: train age_loss mean: 4.293456 2019-10-09 08:14:36.661: INFO @log_variables: train gender_loss mean: 0.041940 2019-10-09 08:14:36.662: INFO @log_variables: train matching_loss nanmean: 0.323991 2019-10-09 08:14:36.662: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:14:36.662: INFO @log_variables: train age_mae mean: 4.768799 2019-10-09 08:14:36.662: INFO @log_variables: train gender_accuracy mean: 0.985439 2019-10-09 08:14:36.662: INFO @log_variables: train positive_distance nanmean: 0.746248 2019-10-09 08:14:36.662: INFO @log_variables: train negative_distance nanmean: 1.408166 2019-10-09 08:14:36.662: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:14:36.662: INFO @log_variables: valid loss mean: 0.454965 2019-10-09 08:14:36.662: INFO @log_variables: valid age_loss mean: 6.359815 2019-10-09 08:14:36.662: INFO @log_variables: valid gender_loss mean: 0.270017 2019-10-09 08:14:36.662: INFO @log_variables: valid matching_loss nanmean: 0.504392 2019-10-09 08:14:36.662: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:14:36.662: INFO @log_variables: valid age_mae mean: 6.842351 2019-10-09 08:14:36.662: INFO @log_variables: valid gender_accuracy mean: 0.925952 2019-10-09 08:14:36.662: INFO @log_variables: valid positive_distance nanmean: 0.779278 2019-10-09 08:14:36.662: INFO @log_variables: valid negative_distance nanmean: 1.371192 2019-10-09 08:14:36.662: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:14:36.662: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:14:39.161: INFO @metrics_hook: valid matching accuracy: 0.8902080709959825, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:14:39.574: INFO @decay_lr : LR updated to `0.00049323554` 2019-10-09 08:14:39.575: INFO @log_profile : T train: 179.717395 2019-10-09 08:14:39.575: INFO @log_profile : T valid: 8.481411 2019-10-09 08:14:39.575: INFO @log_profile : T read data: 1.032164 2019-10-09 08:14:39.576: INFO @log_profile : T hooks: 3.489263 2019-10-09 08:14:39.576: INFO @main_loop : Epoch 141 done 2019-10-09 08:14:39.576: INFO @main_loop : Training epoch 142 2019-10-09 08:17:50.223: INFO @log_variables: train loss mean: 0.257907 2019-10-09 08:17:50.224: INFO @log_variables: train age_loss mean: 4.305512 2019-10-09 08:17:50.224: INFO @log_variables: train gender_loss mean: 0.039563 2019-10-09 08:17:50.224: INFO @log_variables: train matching_loss nanmean: 0.329396 2019-10-09 08:17:50.224: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:17:50.224: INFO @log_variables: train age_mae mean: 4.781017 2019-10-09 08:17:50.224: INFO @log_variables: train gender_accuracy mean: 0.985726 2019-10-09 08:17:50.224: INFO @log_variables: train positive_distance nanmean: 0.748968 2019-10-09 08:17:50.224: INFO @log_variables: train negative_distance nanmean: 1.408344 2019-10-09 08:17:50.224: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:17:50.224: INFO @log_variables: valid loss mean: 0.456700 2019-10-09 08:17:50.224: INFO @log_variables: valid age_loss mean: 6.368620 2019-10-09 08:17:50.224: INFO @log_variables: valid gender_loss mean: 0.280110 2019-10-09 08:17:50.224: INFO @log_variables: valid matching_loss nanmean: 0.498797 2019-10-09 08:17:50.224: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:17:50.224: INFO @log_variables: valid age_mae mean: 6.851202 2019-10-09 08:17:50.224: INFO @log_variables: valid gender_accuracy mean: 0.918855 2019-10-09 08:17:50.224: INFO @log_variables: valid positive_distance nanmean: 0.790565 2019-10-09 08:17:50.224: INFO @log_variables: valid negative_distance nanmean: 1.376451 2019-10-09 08:17:50.225: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:17:50.225: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:17:52.766: INFO @metrics_hook: valid matching accuracy: 0.8905678479342808, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:17:53.216: INFO @decay_lr : LR updated to `0.00049076934` 2019-10-09 08:17:53.217: INFO @log_profile : T train: 180.057023 2019-10-09 08:17:53.217: INFO @log_profile : T valid: 8.551974 2019-10-09 08:17:53.217: INFO @log_profile : T read data: 1.395008 2019-10-09 08:17:53.217: INFO @log_profile : T hooks: 3.553152 2019-10-09 08:17:53.217: INFO @main_loop : Epoch 142 done 2019-10-09 08:17:53.217: INFO @main_loop : Training epoch 143 2019-10-09 08:21:03.591: INFO @log_variables: train loss mean: 0.255627 2019-10-09 08:21:03.591: INFO @log_variables: train age_loss mean: 4.282182 2019-10-09 08:21:03.591: INFO @log_variables: train gender_loss mean: 0.038986 2019-10-09 08:21:03.591: INFO @log_variables: train matching_loss nanmean: 0.325241 2019-10-09 08:21:03.591: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:21:03.592: INFO @log_variables: train age_mae mean: 4.757576 2019-10-09 08:21:03.592: INFO @log_variables: train gender_accuracy mean: 0.986496 2019-10-09 08:21:03.592: INFO @log_variables: train positive_distance nanmean: 0.747427 2019-10-09 08:21:03.592: INFO @log_variables: train negative_distance nanmean: 1.408510 2019-10-09 08:21:03.592: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:21:03.592: INFO @log_variables: valid loss mean: 0.450255 2019-10-09 08:21:03.592: INFO @log_variables: valid age_loss mean: 6.403708 2019-10-09 08:21:03.592: INFO @log_variables: valid gender_loss mean: 0.256082 2019-10-09 08:21:03.592: INFO @log_variables: valid matching_loss nanmean: 0.499337 2019-10-09 08:21:03.592: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:21:03.592: INFO @log_variables: valid age_mae mean: 6.887354 2019-10-09 08:21:03.592: INFO @log_variables: valid gender_accuracy mean: 0.923823 2019-10-09 08:21:03.592: INFO @log_variables: valid positive_distance nanmean: 0.785812 2019-10-09 08:21:03.592: INFO @log_variables: valid negative_distance nanmean: 1.376616 2019-10-09 08:21:03.592: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:21:03.592: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:21:05.771: INFO @metrics_hook: valid matching accuracy: 0.8882892606583918, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:21:06.221: INFO @decay_lr : LR updated to `0.0004883155` 2019-10-09 08:21:06.222: INFO @log_profile : T train: 179.799867 2019-10-09 08:21:06.222: INFO @log_profile : T valid: 8.501545 2019-10-09 08:21:06.222: INFO @log_profile : T read data: 1.426618 2019-10-09 08:21:06.222: INFO @log_profile : T hooks: 3.191377 2019-10-09 08:21:06.222: INFO @main_loop : Epoch 143 done 2019-10-09 08:21:06.222: INFO @main_loop : Training epoch 144 2019-10-09 08:24:16.178: INFO @log_variables: train loss mean: 0.257034 2019-10-09 08:24:16.178: INFO @log_variables: train age_loss mean: 4.314410 2019-10-09 08:24:16.178: INFO @log_variables: train gender_loss mean: 0.038945 2019-10-09 08:24:16.178: INFO @log_variables: train matching_loss nanmean: 0.326419 2019-10-09 08:24:16.178: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:24:16.178: INFO @log_variables: train age_mae mean: 4.790085 2019-10-09 08:24:16.178: INFO @log_variables: train gender_accuracy mean: 0.986152 2019-10-09 08:24:16.178: INFO @log_variables: train positive_distance nanmean: 0.747146 2019-10-09 08:24:16.179: INFO @log_variables: train negative_distance nanmean: 1.408411 2019-10-09 08:24:16.179: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:24:16.179: INFO @log_variables: valid loss mean: 0.453324 2019-10-09 08:24:16.179: INFO @log_variables: valid age_loss mean: 6.275910 2019-10-09 08:24:16.179: INFO @log_variables: valid gender_loss mean: 0.278909 2019-10-09 08:24:16.179: INFO @log_variables: valid matching_loss nanmean: 0.498803 2019-10-09 08:24:16.179: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:24:16.179: INFO @log_variables: valid age_mae mean: 6.758803 2019-10-09 08:24:16.179: INFO @log_variables: valid gender_accuracy mean: 0.921162 2019-10-09 08:24:16.179: INFO @log_variables: valid positive_distance nanmean: 0.788423 2019-10-09 08:24:16.179: INFO @log_variables: valid negative_distance nanmean: 1.375930 2019-10-09 08:24:16.179: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:24:16.179: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:24:18.500: INFO @metrics_hook: valid matching accuracy: 0.8912874018108773, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:24:18.944: INFO @decay_lr : LR updated to `0.0004858739` 2019-10-09 08:24:18.945: INFO @log_profile : T train: 179.782633 2019-10-09 08:24:18.945: INFO @log_profile : T valid: 8.542149 2019-10-09 08:24:18.945: INFO @log_profile : T read data: 0.963414 2019-10-09 08:24:18.945: INFO @log_profile : T hooks: 3.349728 2019-10-09 08:24:18.945: INFO @main_loop : Epoch 144 done 2019-10-09 08:24:18.945: INFO @main_loop : Training epoch 145 2019-10-09 08:27:29.285: INFO @log_variables: train loss mean: 0.254390 2019-10-09 08:27:29.285: INFO @log_variables: train age_loss mean: 4.277690 2019-10-09 08:27:29.285: INFO @log_variables: train gender_loss mean: 0.037615 2019-10-09 08:27:29.285: INFO @log_variables: train matching_loss nanmean: 0.323225 2019-10-09 08:27:29.285: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:27:29.285: INFO @log_variables: train age_mae mean: 4.753053 2019-10-09 08:27:29.286: INFO @log_variables: train gender_accuracy mean: 0.986823 2019-10-09 08:27:29.286: INFO @log_variables: train positive_distance nanmean: 0.745556 2019-10-09 08:27:29.286: INFO @log_variables: train negative_distance nanmean: 1.408354 2019-10-09 08:27:29.286: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:27:29.286: INFO @log_variables: valid loss mean: 0.457589 2019-10-09 08:27:29.286: INFO @log_variables: valid age_loss mean: 6.445051 2019-10-09 08:27:29.286: INFO @log_variables: valid gender_loss mean: 0.268670 2019-10-09 08:27:29.286: INFO @log_variables: valid matching_loss nanmean: 0.505352 2019-10-09 08:27:29.286: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:27:29.286: INFO @log_variables: valid age_mae mean: 6.927525 2019-10-09 08:27:29.286: INFO @log_variables: valid gender_accuracy mean: 0.924060 2019-10-09 08:27:29.286: INFO @log_variables: valid positive_distance nanmean: 0.785297 2019-10-09 08:27:29.286: INFO @log_variables: valid negative_distance nanmean: 1.373460 2019-10-09 08:27:29.286: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:27:29.286: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:27:31.662: INFO @metrics_hook: valid matching accuracy: 0.8891287401810877, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:27:32.116: INFO @decay_lr : LR updated to `0.00048344454` 2019-10-09 08:27:32.118: INFO @log_profile : T train: 179.712993 2019-10-09 08:27:32.118: INFO @log_profile : T valid: 8.561550 2019-10-09 08:27:32.118: INFO @log_profile : T read data: 1.413929 2019-10-09 08:27:32.118: INFO @log_profile : T hooks: 3.398196 2019-10-09 08:27:32.118: INFO @main_loop : Epoch 145 done 2019-10-09 08:27:32.118: INFO @main_loop : Training epoch 146 2019-10-09 08:30:42.331: INFO @log_variables: train loss mean: 0.257025 2019-10-09 08:30:42.331: INFO @log_variables: train age_loss mean: 4.287164 2019-10-09 08:30:42.331: INFO @log_variables: train gender_loss mean: 0.038971 2019-10-09 08:30:42.331: INFO @log_variables: train matching_loss nanmean: 0.329091 2019-10-09 08:30:42.331: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:30:42.331: INFO @log_variables: train age_mae mean: 4.762511 2019-10-09 08:30:42.332: INFO @log_variables: train gender_accuracy mean: 0.986170 2019-10-09 08:30:42.332: INFO @log_variables: train positive_distance nanmean: 0.748257 2019-10-09 08:30:42.332: INFO @log_variables: train negative_distance nanmean: 1.408418 2019-10-09 08:30:42.332: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:30:42.332: INFO @log_variables: valid loss mean: 0.451140 2019-10-09 08:30:42.332: INFO @log_variables: valid age_loss mean: 6.310750 2019-10-09 08:30:42.332: INFO @log_variables: valid gender_loss mean: 0.267090 2019-10-09 08:30:42.332: INFO @log_variables: valid matching_loss nanmean: 0.500370 2019-10-09 08:30:42.332: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:30:42.332: INFO @log_variables: valid age_mae mean: 6.792973 2019-10-09 08:30:42.332: INFO @log_variables: valid gender_accuracy mean: 0.920984 2019-10-09 08:30:42.332: INFO @log_variables: valid positive_distance nanmean: 0.786031 2019-10-09 08:30:42.332: INFO @log_variables: valid negative_distance nanmean: 1.376296 2019-10-09 08:30:42.332: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:30:42.332: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:30:44.420: INFO @metrics_hook: valid matching accuracy: 0.8880494093661929, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:30:44.861: INFO @decay_lr : LR updated to `0.00048102732` 2019-10-09 08:30:44.862: INFO @log_profile : T train: 179.949597 2019-10-09 08:30:44.862: INFO @log_profile : T valid: 8.591837 2019-10-09 08:30:44.862: INFO @log_profile : T read data: 0.993361 2019-10-09 08:30:44.862: INFO @log_profile : T hooks: 3.124035 2019-10-09 08:30:44.863: INFO @main_loop : Epoch 146 done 2019-10-09 08:30:44.863: INFO @main_loop : Training epoch 147 2019-10-09 08:33:55.411: INFO @log_variables: train loss mean: 0.255383 2019-10-09 08:33:55.411: INFO @log_variables: train age_loss mean: 4.267823 2019-10-09 08:33:55.411: INFO @log_variables: train gender_loss mean: 0.038566 2019-10-09 08:33:55.411: INFO @log_variables: train matching_loss nanmean: 0.326340 2019-10-09 08:33:55.411: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:33:55.411: INFO @log_variables: train age_mae mean: 4.742992 2019-10-09 08:33:55.411: INFO @log_variables: train gender_accuracy mean: 0.986430 2019-10-09 08:33:55.411: INFO @log_variables: train positive_distance nanmean: 0.747472 2019-10-09 08:33:55.411: INFO @log_variables: train negative_distance nanmean: 1.408227 2019-10-09 08:33:55.411: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:33:55.411: INFO @log_variables: valid loss mean: 0.452256 2019-10-09 08:33:55.411: INFO @log_variables: valid age_loss mean: 6.349422 2019-10-09 08:33:55.412: INFO @log_variables: valid gender_loss mean: 0.266965 2019-10-09 08:33:55.412: INFO @log_variables: valid matching_loss nanmean: 0.500086 2019-10-09 08:33:55.412: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:33:55.412: INFO @log_variables: valid age_mae mean: 6.832578 2019-10-09 08:33:55.412: INFO @log_variables: valid gender_accuracy mean: 0.923941 2019-10-09 08:33:55.412: INFO @log_variables: valid positive_distance nanmean: 0.787494 2019-10-09 08:33:55.412: INFO @log_variables: valid negative_distance nanmean: 1.377530 2019-10-09 08:33:55.412: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:33:55.412: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:33:57.652: INFO @metrics_hook: valid matching accuracy: 0.8884691491275409, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:33:58.097: INFO @decay_lr : LR updated to `0.00047862218` 2019-10-09 08:33:58.098: INFO @log_profile : T train: 179.911063 2019-10-09 08:33:58.098: INFO @log_profile : T valid: 8.537519 2019-10-09 08:33:58.098: INFO @log_profile : T read data: 1.443154 2019-10-09 08:33:58.098: INFO @log_profile : T hooks: 3.256593 2019-10-09 08:33:58.098: INFO @main_loop : Epoch 147 done 2019-10-09 08:33:58.098: INFO @main_loop : Training epoch 148 2019-10-09 08:37:08.814: INFO @log_variables: train loss mean: 0.255641 2019-10-09 08:37:08.814: INFO @log_variables: train age_loss mean: 4.286157 2019-10-09 08:37:08.814: INFO @log_variables: train gender_loss mean: 0.037232 2019-10-09 08:37:08.814: INFO @log_variables: train matching_loss nanmean: 0.326640 2019-10-09 08:37:08.814: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:37:08.814: INFO @log_variables: train age_mae mean: 4.761680 2019-10-09 08:37:08.814: INFO @log_variables: train gender_accuracy mean: 0.986938 2019-10-09 08:37:08.814: INFO @log_variables: train positive_distance nanmean: 0.748681 2019-10-09 08:37:08.814: INFO @log_variables: train negative_distance nanmean: 1.408246 2019-10-09 08:37:08.814: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:37:08.814: INFO @log_variables: valid loss mean: 0.454749 2019-10-09 08:37:08.814: INFO @log_variables: valid age_loss mean: 6.397108 2019-10-09 08:37:08.814: INFO @log_variables: valid gender_loss mean: 0.273524 2019-10-09 08:37:08.814: INFO @log_variables: valid matching_loss nanmean: 0.496488 2019-10-09 08:37:08.814: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:37:08.815: INFO @log_variables: valid age_mae mean: 6.879897 2019-10-09 08:37:08.815: INFO @log_variables: valid gender_accuracy mean: 0.922877 2019-10-09 08:37:08.815: INFO @log_variables: valid positive_distance nanmean: 0.789662 2019-10-09 08:37:08.815: INFO @log_variables: valid negative_distance nanmean: 1.377255 2019-10-09 08:37:08.815: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:37:08.815: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:37:10.836: INFO @metrics_hook: valid matching accuracy: 0.8894285542963363, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:37:11.270: INFO @decay_lr : LR updated to `0.00047622906` 2019-10-09 08:37:11.271: INFO @log_profile : T train: 180.146027 2019-10-09 08:37:11.271: INFO @log_profile : T valid: 8.505865 2019-10-09 08:37:11.271: INFO @log_profile : T read data: 1.390629 2019-10-09 08:37:11.272: INFO @log_profile : T hooks: 3.044420 2019-10-09 08:37:11.272: INFO @main_loop : Epoch 148 done 2019-10-09 08:37:11.272: INFO @main_loop : Training epoch 149 2019-10-09 08:40:20.863: INFO @log_variables: train loss mean: 0.256684 2019-10-09 08:40:20.864: INFO @log_variables: train age_loss mean: 4.304938 2019-10-09 08:40:20.864: INFO @log_variables: train gender_loss mean: 0.038545 2019-10-09 08:40:20.864: INFO @log_variables: train matching_loss nanmean: 0.326681 2019-10-09 08:40:20.864: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:40:20.864: INFO @log_variables: train age_mae mean: 4.780208 2019-10-09 08:40:20.864: INFO @log_variables: train gender_accuracy mean: 0.985952 2019-10-09 08:40:20.864: INFO @log_variables: train positive_distance nanmean: 0.746597 2019-10-09 08:40:20.864: INFO @log_variables: train negative_distance nanmean: 1.408171 2019-10-09 08:40:20.864: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:40:20.864: INFO @log_variables: valid loss mean: 0.463520 2019-10-09 08:40:20.864: INFO @log_variables: valid age_loss mean: 6.510426 2019-10-09 08:40:20.864: INFO @log_variables: valid gender_loss mean: 0.287582 2019-10-09 08:40:20.864: INFO @log_variables: valid matching_loss nanmean: 0.498289 2019-10-09 08:40:20.864: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:40:20.864: INFO @log_variables: valid age_mae mean: 6.994869 2019-10-09 08:40:20.864: INFO @log_variables: valid gender_accuracy mean: 0.920748 2019-10-09 08:40:20.864: INFO @log_variables: valid positive_distance nanmean: 0.790292 2019-10-09 08:40:20.865: INFO @log_variables: valid negative_distance nanmean: 1.377594 2019-10-09 08:40:20.865: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:40:20.865: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:40:22.998: INFO @metrics_hook: valid matching accuracy: 0.8905678479342808, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:40:23.470: INFO @decay_lr : LR updated to `0.0004738479` 2019-10-09 08:40:23.471: INFO @log_profile : T train: 179.466872 2019-10-09 08:40:23.472: INFO @log_profile : T valid: 8.489308 2019-10-09 08:40:23.472: INFO @log_profile : T read data: 0.978636 2019-10-09 08:40:23.472: INFO @log_profile : T hooks: 3.179288 2019-10-09 08:40:23.472: INFO @main_loop : Epoch 149 done 2019-10-09 08:40:23.472: INFO @main_loop : Training epoch 150 2019-10-09 08:43:33.997: INFO @log_variables: train loss mean: 0.254373 2019-10-09 08:43:33.997: INFO @log_variables: train age_loss mean: 4.266840 2019-10-09 08:43:33.997: INFO @log_variables: train gender_loss mean: 0.036765 2019-10-09 08:43:33.997: INFO @log_variables: train matching_loss nanmean: 0.325106 2019-10-09 08:43:33.997: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:43:33.997: INFO @log_variables: train age_mae mean: 4.741728 2019-10-09 08:43:33.997: INFO @log_variables: train gender_accuracy mean: 0.986633 2019-10-09 08:43:33.997: INFO @log_variables: train positive_distance nanmean: 0.746208 2019-10-09 08:43:33.997: INFO @log_variables: train negative_distance nanmean: 1.407981 2019-10-09 08:43:33.997: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:43:33.997: INFO @log_variables: valid loss mean: 0.456019 2019-10-09 08:43:33.997: INFO @log_variables: valid age_loss mean: 6.579241 2019-10-09 08:43:33.997: INFO @log_variables: valid gender_loss mean: 0.255398 2019-10-09 08:43:33.997: INFO @log_variables: valid matching_loss nanmean: 0.500337 2019-10-09 08:43:33.997: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:43:33.997: INFO @log_variables: valid age_mae mean: 7.062686 2019-10-09 08:43:33.998: INFO @log_variables: valid gender_accuracy mean: 0.924769 2019-10-09 08:43:33.998: INFO @log_variables: valid positive_distance nanmean: 0.783480 2019-10-09 08:43:33.998: INFO @log_variables: valid negative_distance nanmean: 1.374764 2019-10-09 08:43:33.998: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:43:33.998: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:43:36.563: INFO @metrics_hook: valid matching accuracy: 0.8887689632427894, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:43:36.997: INFO @decay_lr : LR updated to `0.00047147868` 2019-10-09 08:43:36.998: INFO @log_profile : T train: 179.851236 2019-10-09 08:43:36.998: INFO @log_profile : T valid: 8.585251 2019-10-09 08:43:36.998: INFO @log_profile : T read data: 1.416318 2019-10-09 08:43:36.998: INFO @log_profile : T hooks: 3.587684 2019-10-09 08:43:36.998: INFO @main_loop : Epoch 150 done 2019-10-09 08:43:36.998: INFO @main_loop : Training epoch 151 2019-10-09 08:46:47.087: INFO @log_variables: train loss mean: 0.253643 2019-10-09 08:46:47.087: INFO @log_variables: train age_loss mean: 4.273711 2019-10-09 08:46:47.087: INFO @log_variables: train gender_loss mean: 0.036697 2019-10-09 08:46:47.088: INFO @log_variables: train matching_loss nanmean: 0.322225 2019-10-09 08:46:47.088: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:46:47.088: INFO @log_variables: train age_mae mean: 4.749014 2019-10-09 08:46:47.088: INFO @log_variables: train gender_accuracy mean: 0.986757 2019-10-09 08:46:47.088: INFO @log_variables: train positive_distance nanmean: 0.744941 2019-10-09 08:46:47.088: INFO @log_variables: train negative_distance nanmean: 1.408291 2019-10-09 08:46:47.088: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:46:47.088: INFO @log_variables: valid loss mean: 0.457560 2019-10-09 08:46:47.088: INFO @log_variables: valid age_loss mean: 6.486586 2019-10-09 08:46:47.088: INFO @log_variables: valid gender_loss mean: 0.274239 2019-10-09 08:46:47.088: INFO @log_variables: valid matching_loss nanmean: 0.495539 2019-10-09 08:46:47.088: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:46:47.088: INFO @log_variables: valid age_mae mean: 6.969851 2019-10-09 08:46:47.088: INFO @log_variables: valid gender_accuracy mean: 0.926130 2019-10-09 08:46:47.088: INFO @log_variables: valid positive_distance nanmean: 0.789773 2019-10-09 08:46:47.088: INFO @log_variables: valid negative_distance nanmean: 1.377194 2019-10-09 08:46:47.088: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:46:47.089: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:46:49.471: INFO @metrics_hook: valid matching accuracy: 0.8881093721892427, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:46:49.915: INFO @decay_lr : LR updated to `0.0004691213` 2019-10-09 08:46:49.916: INFO @log_profile : T train: 179.856838 2019-10-09 08:46:49.916: INFO @log_profile : T valid: 8.568099 2019-10-09 08:46:49.916: INFO @log_profile : T read data: 1.008603 2019-10-09 08:46:49.916: INFO @log_profile : T hooks: 3.399578 2019-10-09 08:46:49.916: INFO @main_loop : Epoch 151 done 2019-10-09 08:46:49.916: INFO @main_loop : Training epoch 152 2019-10-09 08:50:00.130: INFO @log_variables: train loss mean: 0.254384 2019-10-09 08:50:00.131: INFO @log_variables: train age_loss mean: 4.251067 2019-10-09 08:50:00.131: INFO @log_variables: train gender_loss mean: 0.037847 2019-10-09 08:50:00.131: INFO @log_variables: train matching_loss nanmean: 0.325637 2019-10-09 08:50:00.131: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:50:00.131: INFO @log_variables: train age_mae mean: 4.726776 2019-10-09 08:50:00.131: INFO @log_variables: train gender_accuracy mean: 0.986649 2019-10-09 08:50:00.131: INFO @log_variables: train positive_distance nanmean: 0.746047 2019-10-09 08:50:00.131: INFO @log_variables: train negative_distance nanmean: 1.408088 2019-10-09 08:50:00.131: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:50:00.131: INFO @log_variables: valid loss mean: 0.444331 2019-10-09 08:50:00.131: INFO @log_variables: valid age_loss mean: 6.254659 2019-10-09 08:50:00.131: INFO @log_variables: valid gender_loss mean: 0.259444 2019-10-09 08:50:00.131: INFO @log_variables: valid matching_loss nanmean: 0.492517 2019-10-09 08:50:00.131: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:50:00.131: INFO @log_variables: valid age_mae mean: 6.737264 2019-10-09 08:50:00.131: INFO @log_variables: valid gender_accuracy mean: 0.922818 2019-10-09 08:50:00.132: INFO @log_variables: valid positive_distance nanmean: 0.784731 2019-10-09 08:50:00.132: INFO @log_variables: valid negative_distance nanmean: 1.375938 2019-10-09 08:50:00.132: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:50:00.132: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:50:02.452: INFO @metrics_hook: valid matching accuracy: 0.88864903759669, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:50:02.887: INFO @decay_lr : LR updated to `0.0004667757` 2019-10-09 08:50:02.888: INFO @log_profile : T train: 179.665854 2019-10-09 08:50:02.888: INFO @log_profile : T valid: 8.522835 2019-10-09 08:50:02.888: INFO @log_profile : T read data: 1.353403 2019-10-09 08:50:02.888: INFO @log_profile : T hooks: 3.343968 2019-10-09 08:50:02.888: INFO @main_loop : Epoch 152 done 2019-10-09 08:50:02.889: INFO @main_loop : Training epoch 153 2019-10-09 08:53:13.607: INFO @log_variables: train loss mean: 0.252759 2019-10-09 08:53:13.607: INFO @log_variables: train age_loss mean: 4.245569 2019-10-09 08:53:13.608: INFO @log_variables: train gender_loss mean: 0.038351 2019-10-09 08:53:13.608: INFO @log_variables: train matching_loss nanmean: 0.320645 2019-10-09 08:53:13.608: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:53:13.608: INFO @log_variables: train age_mae mean: 4.720881 2019-10-09 08:53:13.608: INFO @log_variables: train gender_accuracy mean: 0.986286 2019-10-09 08:53:13.608: INFO @log_variables: train positive_distance nanmean: 0.743578 2019-10-09 08:53:13.608: INFO @log_variables: train negative_distance nanmean: 1.408294 2019-10-09 08:53:13.608: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:53:13.608: INFO @log_variables: valid loss mean: 0.469408 2019-10-09 08:53:13.608: INFO @log_variables: valid age_loss mean: 6.509613 2019-10-09 08:53:13.608: INFO @log_variables: valid gender_loss mean: 0.304146 2019-10-09 08:53:13.608: INFO @log_variables: valid matching_loss nanmean: 0.500059 2019-10-09 08:53:13.608: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:53:13.608: INFO @log_variables: valid age_mae mean: 6.992865 2019-10-09 08:53:13.608: INFO @log_variables: valid gender_accuracy mean: 0.917317 2019-10-09 08:53:13.608: INFO @log_variables: valid positive_distance nanmean: 0.786914 2019-10-09 08:53:13.608: INFO @log_variables: valid negative_distance nanmean: 1.372274 2019-10-09 08:53:13.608: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:53:13.608: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:53:15.786: INFO @metrics_hook: valid matching accuracy: 0.8877495952509444, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:53:16.230: INFO @decay_lr : LR updated to `0.00046444184` 2019-10-09 08:53:16.231: INFO @log_profile : T train: 180.024539 2019-10-09 08:53:16.231: INFO @log_profile : T valid: 8.545759 2019-10-09 08:53:16.231: INFO @log_profile : T read data: 1.481755 2019-10-09 08:53:16.232: INFO @log_profile : T hooks: 3.205967 2019-10-09 08:53:16.232: INFO @main_loop : Epoch 153 done 2019-10-09 08:53:16.232: INFO @main_loop : Training epoch 154 2019-10-09 08:56:26.348: INFO @log_variables: train loss mean: 0.254455 2019-10-09 08:56:26.348: INFO @log_variables: train age_loss mean: 4.251787 2019-10-09 08:56:26.348: INFO @log_variables: train gender_loss mean: 0.040265 2019-10-09 08:56:26.348: INFO @log_variables: train matching_loss nanmean: 0.323368 2019-10-09 08:56:26.348: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:56:26.348: INFO @log_variables: train age_mae mean: 4.726969 2019-10-09 08:56:26.348: INFO @log_variables: train gender_accuracy mean: 0.985628 2019-10-09 08:56:26.348: INFO @log_variables: train positive_distance nanmean: 0.747303 2019-10-09 08:56:26.348: INFO @log_variables: train negative_distance nanmean: 1.408014 2019-10-09 08:56:26.348: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:56:26.348: INFO @log_variables: valid loss mean: 0.451293 2019-10-09 08:56:26.349: INFO @log_variables: valid age_loss mean: 6.528470 2019-10-09 08:56:26.349: INFO @log_variables: valid gender_loss mean: 0.246003 2019-10-09 08:56:26.349: INFO @log_variables: valid matching_loss nanmean: 0.500160 2019-10-09 08:56:26.349: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:56:26.349: INFO @log_variables: valid age_mae mean: 7.011563 2019-10-09 08:56:26.349: INFO @log_variables: valid gender_accuracy mean: 0.928495 2019-10-09 08:56:26.349: INFO @log_variables: valid positive_distance nanmean: 0.784529 2019-10-09 08:56:26.349: INFO @log_variables: valid negative_distance nanmean: 1.373717 2019-10-09 08:56:26.349: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:56:26.349: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:56:28.527: INFO @metrics_hook: valid matching accuracy: 0.8880494093661929, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:56:28.958: INFO @decay_lr : LR updated to `0.00046211964` 2019-10-09 08:56:28.960: INFO @log_profile : T train: 179.884237 2019-10-09 08:56:28.960: INFO @log_profile : T valid: 8.543688 2019-10-09 08:56:28.960: INFO @log_profile : T read data: 1.011464 2019-10-09 08:56:28.960: INFO @log_profile : T hooks: 3.200250 2019-10-09 08:56:28.960: INFO @main_loop : Epoch 154 done 2019-10-09 08:56:28.960: INFO @main_loop : Training epoch 155 2019-10-09 08:59:39.376: INFO @log_variables: train loss mean: 0.253076 2019-10-09 08:59:39.376: INFO @log_variables: train age_loss mean: 4.219110 2019-10-09 08:59:39.376: INFO @log_variables: train gender_loss mean: 0.039060 2019-10-09 08:59:39.376: INFO @log_variables: train matching_loss nanmean: 0.323566 2019-10-09 08:59:39.376: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 08:59:39.376: INFO @log_variables: train age_mae mean: 4.693912 2019-10-09 08:59:39.376: INFO @log_variables: train gender_accuracy mean: 0.986350 2019-10-09 08:59:39.376: INFO @log_variables: train positive_distance nanmean: 0.745581 2019-10-09 08:59:39.376: INFO @log_variables: train negative_distance nanmean: 1.407982 2019-10-09 08:59:39.376: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 08:59:39.376: INFO @log_variables: valid loss mean: 0.448604 2019-10-09 08:59:39.376: INFO @log_variables: valid age_loss mean: 6.429662 2019-10-09 08:59:39.376: INFO @log_variables: valid gender_loss mean: 0.250445 2019-10-09 08:59:39.376: INFO @log_variables: valid matching_loss nanmean: 0.497260 2019-10-09 08:59:39.377: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 08:59:39.377: INFO @log_variables: valid age_mae mean: 6.913737 2019-10-09 08:59:39.377: INFO @log_variables: valid gender_accuracy mean: 0.927549 2019-10-09 08:59:39.377: INFO @log_variables: valid positive_distance nanmean: 0.791440 2019-10-09 08:59:39.377: INFO @log_variables: valid negative_distance nanmean: 1.377551 2019-10-09 08:59:39.377: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 08:59:39.377: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 08:59:41.780: INFO @metrics_hook: valid matching accuracy: 0.8907477364034299, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 08:59:42.292: INFO @decay_lr : LR updated to `0.00045980903` 2019-10-09 08:59:42.294: INFO @log_profile : T train: 179.707848 2019-10-09 08:59:42.294: INFO @log_profile : T valid: 8.566967 2019-10-09 08:59:42.294: INFO @log_profile : T read data: 1.479748 2019-10-09 08:59:42.294: INFO @log_profile : T hooks: 3.491928 2019-10-09 08:59:42.294: INFO @main_loop : Epoch 155 done 2019-10-09 08:59:42.294: INFO @main_loop : Training epoch 156 2019-10-09 09:02:52.700: INFO @log_variables: train loss mean: 0.254914 2019-10-09 09:02:52.700: INFO @log_variables: train age_loss mean: 4.256253 2019-10-09 09:02:52.700: INFO @log_variables: train gender_loss mean: 0.039520 2019-10-09 09:02:52.700: INFO @log_variables: train matching_loss nanmean: 0.325089 2019-10-09 09:02:52.700: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:02:52.701: INFO @log_variables: train age_mae mean: 4.731322 2019-10-09 09:02:52.701: INFO @log_variables: train gender_accuracy mean: 0.986542 2019-10-09 09:02:52.701: INFO @log_variables: train positive_distance nanmean: 0.745550 2019-10-09 09:02:52.701: INFO @log_variables: train negative_distance nanmean: 1.408001 2019-10-09 09:02:52.701: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:02:52.701: INFO @log_variables: valid loss mean: 0.461643 2019-10-09 09:02:52.701: INFO @log_variables: valid age_loss mean: 6.538265 2019-10-09 09:02:52.701: INFO @log_variables: valid gender_loss mean: 0.281970 2019-10-09 09:02:52.701: INFO @log_variables: valid matching_loss nanmean: 0.495298 2019-10-09 09:02:52.701: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:02:52.701: INFO @log_variables: valid age_mae mean: 7.021444 2019-10-09 09:02:52.701: INFO @log_variables: valid gender_accuracy mean: 0.921812 2019-10-09 09:02:52.701: INFO @log_variables: valid positive_distance nanmean: 0.789017 2019-10-09 09:02:52.701: INFO @log_variables: valid negative_distance nanmean: 1.379250 2019-10-09 09:02:52.701: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:02:52.701: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:02:55.059: INFO @metrics_hook: valid matching accuracy: 0.8908076992264796, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:02:55.530: INFO @decay_lr : LR updated to `0.00045750997` 2019-10-09 09:02:55.532: INFO @log_profile : T train: 179.762599 2019-10-09 09:02:55.532: INFO @log_profile : T valid: 8.535721 2019-10-09 09:02:55.532: INFO @log_profile : T read data: 1.440234 2019-10-09 09:02:55.532: INFO @log_profile : T hooks: 3.413049 2019-10-09 09:02:55.532: INFO @main_loop : Epoch 156 done 2019-10-09 09:02:55.532: INFO @main_loop : Training epoch 157 2019-10-09 09:06:05.778: INFO @log_variables: train loss mean: 0.251911 2019-10-09 09:06:05.778: INFO @log_variables: train age_loss mean: 4.211647 2019-10-09 09:06:05.778: INFO @log_variables: train gender_loss mean: 0.036693 2019-10-09 09:06:05.778: INFO @log_variables: train matching_loss nanmean: 0.323067 2019-10-09 09:06:05.778: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:06:05.778: INFO @log_variables: train age_mae mean: 4.687033 2019-10-09 09:06:05.778: INFO @log_variables: train gender_accuracy mean: 0.986822 2019-10-09 09:06:05.779: INFO @log_variables: train positive_distance nanmean: 0.746013 2019-10-09 09:06:05.779: INFO @log_variables: train negative_distance nanmean: 1.408202 2019-10-09 09:06:05.779: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:06:05.779: INFO @log_variables: valid loss mean: 0.461026 2019-10-09 09:06:05.779: INFO @log_variables: valid age_loss mean: 6.374295 2019-10-09 09:06:05.779: INFO @log_variables: valid gender_loss mean: 0.295089 2019-10-09 09:06:05.779: INFO @log_variables: valid matching_loss nanmean: 0.496661 2019-10-09 09:06:05.779: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:06:05.779: INFO @log_variables: valid age_mae mean: 6.857187 2019-10-09 09:06:05.779: INFO @log_variables: valid gender_accuracy mean: 0.922285 2019-10-09 09:06:05.779: INFO @log_variables: valid positive_distance nanmean: 0.788946 2019-10-09 09:06:05.779: INFO @log_variables: valid negative_distance nanmean: 1.375682 2019-10-09 09:06:05.779: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:06:05.779: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:06:07.823: INFO @metrics_hook: valid matching accuracy: 0.8891887030041374, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:06:08.264: INFO @decay_lr : LR updated to `0.0004552224` 2019-10-09 09:06:08.265: INFO @log_profile : T train: 180.075531 2019-10-09 09:06:08.265: INFO @log_profile : T valid: 8.496425 2019-10-09 09:06:08.265: INFO @log_profile : T read data: 1.032268 2019-10-09 09:06:08.265: INFO @log_profile : T hooks: 3.043283 2019-10-09 09:06:08.265: INFO @main_loop : Epoch 157 done 2019-10-09 09:06:08.265: INFO @main_loop : Training epoch 158 2019-10-09 09:09:18.988: INFO @log_variables: train loss mean: 0.251722 2019-10-09 09:09:18.988: INFO @log_variables: train age_loss mean: 4.213106 2019-10-09 09:09:18.988: INFO @log_variables: train gender_loss mean: 0.036977 2019-10-09 09:09:18.988: INFO @log_variables: train matching_loss nanmean: 0.322049 2019-10-09 09:09:18.988: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:09:18.988: INFO @log_variables: train age_mae mean: 4.688098 2019-10-09 09:09:18.988: INFO @log_variables: train gender_accuracy mean: 0.987101 2019-10-09 09:09:18.988: INFO @log_variables: train positive_distance nanmean: 0.745246 2019-10-09 09:09:18.988: INFO @log_variables: train negative_distance nanmean: 1.408345 2019-10-09 09:09:18.988: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:09:18.988: INFO @log_variables: valid loss mean: 0.457162 2019-10-09 09:09:18.989: INFO @log_variables: valid age_loss mean: 6.309742 2019-10-09 09:09:18.989: INFO @log_variables: valid gender_loss mean: 0.288667 2019-10-09 09:09:18.989: INFO @log_variables: valid matching_loss nanmean: 0.497561 2019-10-09 09:09:18.989: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:09:18.989: INFO @log_variables: valid age_mae mean: 6.792781 2019-10-09 09:09:18.989: INFO @log_variables: valid gender_accuracy mean: 0.919506 2019-10-09 09:09:18.989: INFO @log_variables: valid positive_distance nanmean: 0.783233 2019-10-09 09:09:18.989: INFO @log_variables: valid negative_distance nanmean: 1.373480 2019-10-09 09:09:18.989: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:09:18.989: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:09:21.368: INFO @metrics_hook: valid matching accuracy: 0.8871499670204473, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:09:21.829: INFO @decay_lr : LR updated to `0.0004529463` 2019-10-09 09:09:21.830: INFO @log_profile : T train: 179.779892 2019-10-09 09:09:21.830: INFO @log_profile : T valid: 8.847575 2019-10-09 09:09:21.830: INFO @log_profile : T read data: 1.437704 2019-10-09 09:09:21.830: INFO @log_profile : T hooks: 3.414316 2019-10-09 09:09:21.830: INFO @main_loop : Epoch 158 done 2019-10-09 09:09:21.830: INFO @main_loop : Training epoch 159 2019-10-09 09:12:31.540: INFO @log_variables: train loss mean: 0.250887 2019-10-09 09:12:31.541: INFO @log_variables: train age_loss mean: 4.231817 2019-10-09 09:12:31.541: INFO @log_variables: train gender_loss mean: 0.035596 2019-10-09 09:12:31.541: INFO @log_variables: train matching_loss nanmean: 0.318974 2019-10-09 09:12:31.541: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:12:31.541: INFO @log_variables: train age_mae mean: 4.707127 2019-10-09 09:12:31.541: INFO @log_variables: train gender_accuracy mean: 0.987500 2019-10-09 09:12:31.541: INFO @log_variables: train positive_distance nanmean: 0.743407 2019-10-09 09:12:31.541: INFO @log_variables: train negative_distance nanmean: 1.408378 2019-10-09 09:12:31.541: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:12:31.541: INFO @log_variables: valid loss mean: 0.454728 2019-10-09 09:12:31.541: INFO @log_variables: valid age_loss mean: 6.407053 2019-10-09 09:12:31.541: INFO @log_variables: valid gender_loss mean: 0.276541 2019-10-09 09:12:31.541: INFO @log_variables: valid matching_loss nanmean: 0.492411 2019-10-09 09:12:31.541: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:12:31.541: INFO @log_variables: valid age_mae mean: 6.890131 2019-10-09 09:12:31.542: INFO @log_variables: valid gender_accuracy mean: 0.923764 2019-10-09 09:12:31.542: INFO @log_variables: valid positive_distance nanmean: 0.788981 2019-10-09 09:12:31.542: INFO @log_variables: valid negative_distance nanmean: 1.379061 2019-10-09 09:12:31.542: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:12:31.542: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:12:34.327: INFO @metrics_hook: valid matching accuracy: 0.8881693350122923, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:12:34.766: INFO @decay_lr : LR updated to `0.00045068155` 2019-10-09 09:12:34.767: INFO @log_profile : T train: 179.560357 2019-10-09 09:12:34.767: INFO @log_profile : T valid: 8.509416 2019-10-09 09:12:34.767: INFO @log_profile : T read data: 1.000154 2019-10-09 09:12:34.768: INFO @log_profile : T hooks: 3.782975 2019-10-09 09:12:34.768: INFO @main_loop : Epoch 159 done 2019-10-09 09:12:34.768: INFO @main_loop : Training epoch 160 2019-10-09 09:15:45.273: INFO @log_variables: train loss mean: 0.251675 2019-10-09 09:15:45.273: INFO @log_variables: train age_loss mean: 4.200350 2019-10-09 09:15:45.274: INFO @log_variables: train gender_loss mean: 0.037754 2019-10-09 09:15:45.274: INFO @log_variables: train matching_loss nanmean: 0.322405 2019-10-09 09:15:45.274: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:15:45.274: INFO @log_variables: train age_mae mean: 4.674931 2019-10-09 09:15:45.274: INFO @log_variables: train gender_accuracy mean: 0.986405 2019-10-09 09:15:45.274: INFO @log_variables: train positive_distance nanmean: 0.745403 2019-10-09 09:15:45.274: INFO @log_variables: train negative_distance nanmean: 1.408049 2019-10-09 09:15:45.274: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:15:45.274: INFO @log_variables: valid loss mean: 0.446802 2019-10-09 09:15:45.274: INFO @log_variables: valid age_loss mean: 6.259249 2019-10-09 09:15:45.274: INFO @log_variables: valid gender_loss mean: 0.262965 2019-10-09 09:15:45.274: INFO @log_variables: valid matching_loss nanmean: 0.496197 2019-10-09 09:15:45.274: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:15:45.274: INFO @log_variables: valid age_mae mean: 6.741193 2019-10-09 09:15:45.274: INFO @log_variables: valid gender_accuracy mean: 0.925656 2019-10-09 09:15:45.274: INFO @log_variables: valid positive_distance nanmean: 0.789456 2019-10-09 09:15:45.274: INFO @log_variables: valid negative_distance nanmean: 1.379920 2019-10-09 09:15:45.274: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:15:45.274: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:15:47.409: INFO @metrics_hook: valid matching accuracy: 0.8875097439587456, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:15:47.845: INFO @decay_lr : LR updated to `0.00044842815` 2019-10-09 09:15:47.847: INFO @log_profile : T train: 179.832327 2019-10-09 09:15:47.847: INFO @log_profile : T valid: 8.530889 2019-10-09 09:15:47.847: INFO @log_profile : T read data: 1.460192 2019-10-09 09:15:47.847: INFO @log_profile : T hooks: 3.169186 2019-10-09 09:15:47.847: INFO @main_loop : Epoch 160 done 2019-10-09 09:15:47.847: INFO @main_loop : Training epoch 161 2019-10-09 09:18:58.355: INFO @log_variables: train loss mean: 0.253578 2019-10-09 09:18:58.355: INFO @log_variables: train age_loss mean: 4.249325 2019-10-09 09:18:58.355: INFO @log_variables: train gender_loss mean: 0.039505 2019-10-09 09:18:58.356: INFO @log_variables: train matching_loss nanmean: 0.321653 2019-10-09 09:18:58.356: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:18:58.356: INFO @log_variables: train age_mae mean: 4.724486 2019-10-09 09:18:58.356: INFO @log_variables: train gender_accuracy mean: 0.986071 2019-10-09 09:18:58.356: INFO @log_variables: train positive_distance nanmean: 0.744881 2019-10-09 09:18:58.356: INFO @log_variables: train negative_distance nanmean: 1.408600 2019-10-09 09:18:58.356: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:18:58.356: INFO @log_variables: valid loss mean: 0.453848 2019-10-09 09:18:58.356: INFO @log_variables: valid age_loss mean: 6.392220 2019-10-09 09:18:58.356: INFO @log_variables: valid gender_loss mean: 0.272602 2019-10-09 09:18:58.356: INFO @log_variables: valid matching_loss nanmean: 0.495104 2019-10-09 09:18:58.356: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:18:58.356: INFO @log_variables: valid age_mae mean: 6.875653 2019-10-09 09:18:58.356: INFO @log_variables: valid gender_accuracy mean: 0.925656 2019-10-09 09:18:58.356: INFO @log_variables: valid positive_distance nanmean: 0.791018 2019-10-09 09:18:58.356: INFO @log_variables: valid negative_distance nanmean: 1.379568 2019-10-09 09:18:58.356: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:18:58.356: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:19:00.834: INFO @metrics_hook: valid matching accuracy: 0.8888289260658392, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:19:01.262: INFO @decay_lr : LR updated to `0.00044618602` 2019-10-09 09:19:01.264: INFO @log_profile : T train: 179.880358 2019-10-09 09:19:01.264: INFO @log_profile : T valid: 8.548814 2019-10-09 09:19:01.264: INFO @log_profile : T read data: 1.407678 2019-10-09 09:19:01.264: INFO @log_profile : T hooks: 3.493820 2019-10-09 09:19:01.264: INFO @main_loop : Epoch 161 done 2019-10-09 09:19:01.264: INFO @main_loop : Training epoch 162 2019-10-09 09:22:11.227: INFO @log_variables: train loss mean: 0.250172 2019-10-09 09:22:11.227: INFO @log_variables: train age_loss mean: 4.189167 2019-10-09 09:22:11.228: INFO @log_variables: train gender_loss mean: 0.037720 2019-10-09 09:22:11.228: INFO @log_variables: train matching_loss nanmean: 0.318896 2019-10-09 09:22:11.228: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:22:11.228: INFO @log_variables: train age_mae mean: 4.664354 2019-10-09 09:22:11.228: INFO @log_variables: train gender_accuracy mean: 0.986522 2019-10-09 09:22:11.228: INFO @log_variables: train positive_distance nanmean: 0.743587 2019-10-09 09:22:11.228: INFO @log_variables: train negative_distance nanmean: 1.408387 2019-10-09 09:22:11.228: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:22:11.228: INFO @log_variables: valid loss mean: 0.459245 2019-10-09 09:22:11.228: INFO @log_variables: valid age_loss mean: 6.338118 2019-10-09 09:22:11.228: INFO @log_variables: valid gender_loss mean: 0.292496 2019-10-09 09:22:11.228: INFO @log_variables: valid matching_loss nanmean: 0.497351 2019-10-09 09:22:11.228: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:22:11.228: INFO @log_variables: valid age_mae mean: 6.822249 2019-10-09 09:22:11.228: INFO @log_variables: valid gender_accuracy mean: 0.921930 2019-10-09 09:22:11.228: INFO @log_variables: valid positive_distance nanmean: 0.785062 2019-10-09 09:22:11.228: INFO @log_variables: valid negative_distance nanmean: 1.376546 2019-10-09 09:22:11.229: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:22:11.229: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:22:13.224: INFO @metrics_hook: valid matching accuracy: 0.8881093721892427, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:22:13.669: INFO @decay_lr : LR updated to `0.0004439551` 2019-10-09 09:22:13.670: INFO @log_profile : T train: 179.732896 2019-10-09 09:22:13.671: INFO @log_profile : T valid: 8.576564 2019-10-09 09:22:13.671: INFO @log_profile : T read data: 0.989445 2019-10-09 09:22:13.671: INFO @log_profile : T hooks: 3.022504 2019-10-09 09:22:13.671: INFO @main_loop : Epoch 162 done 2019-10-09 09:22:13.671: INFO @main_loop : Training epoch 163 2019-10-09 09:25:24.068: INFO @log_variables: train loss mean: 0.251157 2019-10-09 09:25:24.068: INFO @log_variables: train age_loss mean: 4.203657 2019-10-09 09:25:24.069: INFO @log_variables: train gender_loss mean: 0.036977 2019-10-09 09:25:24.069: INFO @log_variables: train matching_loss nanmean: 0.321245 2019-10-09 09:25:24.069: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:25:24.069: INFO @log_variables: train age_mae mean: 4.678613 2019-10-09 09:25:24.069: INFO @log_variables: train gender_accuracy mean: 0.986530 2019-10-09 09:25:24.069: INFO @log_variables: train positive_distance nanmean: 0.744533 2019-10-09 09:25:24.069: INFO @log_variables: train negative_distance nanmean: 1.407920 2019-10-09 09:25:24.069: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:25:24.069: INFO @log_variables: valid loss mean: 0.454714 2019-10-09 09:25:24.069: INFO @log_variables: valid age_loss mean: 6.275976 2019-10-09 09:25:24.069: INFO @log_variables: valid gender_loss mean: 0.284558 2019-10-09 09:25:24.069: INFO @log_variables: valid matching_loss nanmean: 0.497457 2019-10-09 09:25:24.069: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:25:24.069: INFO @log_variables: valid age_mae mean: 6.757026 2019-10-09 09:25:24.069: INFO @log_variables: valid gender_accuracy mean: 0.924178 2019-10-09 09:25:24.069: INFO @log_variables: valid positive_distance nanmean: 0.786885 2019-10-09 09:25:24.069: INFO @log_variables: valid negative_distance nanmean: 1.377847 2019-10-09 09:25:24.069: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:25:24.069: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:25:26.400: INFO @metrics_hook: valid matching accuracy: 0.8897883312346345, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:25:26.849: INFO @decay_lr : LR updated to `0.00044173532` 2019-10-09 09:25:26.851: INFO @log_profile : T train: 179.743319 2019-10-09 09:25:26.851: INFO @log_profile : T valid: 8.552806 2019-10-09 09:25:26.851: INFO @log_profile : T read data: 1.435981 2019-10-09 09:25:26.851: INFO @log_profile : T hooks: 3.361838 2019-10-09 09:25:26.851: INFO @main_loop : Epoch 163 done 2019-10-09 09:25:26.851: INFO @main_loop : Training epoch 164 2019-10-09 09:28:37.072: INFO @log_variables: train loss mean: 0.251710 2019-10-09 09:28:37.072: INFO @log_variables: train age_loss mean: 4.207917 2019-10-09 09:28:37.072: INFO @log_variables: train gender_loss mean: 0.037426 2019-10-09 09:28:37.072: INFO @log_variables: train matching_loss nanmean: 0.322084 2019-10-09 09:28:37.072: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:28:37.072: INFO @log_variables: train age_mae mean: 4.683354 2019-10-09 09:28:37.072: INFO @log_variables: train gender_accuracy mean: 0.986596 2019-10-09 09:28:37.072: INFO @log_variables: train positive_distance nanmean: 0.744938 2019-10-09 09:28:37.072: INFO @log_variables: train negative_distance nanmean: 1.408258 2019-10-09 09:28:37.072: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:28:37.072: INFO @log_variables: valid loss mean: 0.451586 2019-10-09 09:28:37.072: INFO @log_variables: valid age_loss mean: 6.310734 2019-10-09 09:28:37.072: INFO @log_variables: valid gender_loss mean: 0.273128 2019-10-09 09:28:37.073: INFO @log_variables: valid matching_loss nanmean: 0.495714 2019-10-09 09:28:37.073: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:28:37.073: INFO @log_variables: valid age_mae mean: 6.792939 2019-10-09 09:28:37.073: INFO @log_variables: valid gender_accuracy mean: 0.923409 2019-10-09 09:28:37.073: INFO @log_variables: valid positive_distance nanmean: 0.784501 2019-10-09 09:28:37.073: INFO @log_variables: valid negative_distance nanmean: 1.374610 2019-10-09 09:28:37.073: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:28:37.073: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:28:39.601: INFO @metrics_hook: valid matching accuracy: 0.8899082568807339, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:28:40.032: INFO @decay_lr : LR updated to `0.00043952663` 2019-10-09 09:28:40.033: INFO @log_profile : T train: 179.726648 2019-10-09 09:28:40.034: INFO @log_profile : T valid: 8.480364 2019-10-09 09:28:40.034: INFO @log_profile : T read data: 1.363229 2019-10-09 09:28:40.034: INFO @log_profile : T hooks: 3.527956 2019-10-09 09:28:40.034: INFO @main_loop : Epoch 164 done 2019-10-09 09:28:40.034: INFO @main_loop : Training epoch 165 2019-10-09 09:31:50.603: INFO @log_variables: train loss mean: 0.248866 2019-10-09 09:31:50.604: INFO @log_variables: train age_loss mean: 4.188494 2019-10-09 09:31:50.604: INFO @log_variables: train gender_loss mean: 0.033863 2019-10-09 09:31:50.604: INFO @log_variables: train matching_loss nanmean: 0.318771 2019-10-09 09:31:50.604: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:31:50.604: INFO @log_variables: train age_mae mean: 4.663393 2019-10-09 09:31:50.604: INFO @log_variables: train gender_accuracy mean: 0.988114 2019-10-09 09:31:50.604: INFO @log_variables: train positive_distance nanmean: 0.742409 2019-10-09 09:31:50.604: INFO @log_variables: train negative_distance nanmean: 1.408350 2019-10-09 09:31:50.604: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:31:50.604: INFO @log_variables: valid loss mean: 0.453755 2019-10-09 09:31:50.604: INFO @log_variables: valid age_loss mean: 6.435028 2019-10-09 09:31:50.604: INFO @log_variables: valid gender_loss mean: 0.271946 2019-10-09 09:31:50.604: INFO @log_variables: valid matching_loss nanmean: 0.491192 2019-10-09 09:31:50.604: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:31:50.604: INFO @log_variables: valid age_mae mean: 6.917272 2019-10-09 09:31:50.604: INFO @log_variables: valid gender_accuracy mean: 0.926485 2019-10-09 09:31:50.604: INFO @log_variables: valid positive_distance nanmean: 0.790737 2019-10-09 09:31:50.604: INFO @log_variables: valid negative_distance nanmean: 1.376250 2019-10-09 09:31:50.605: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:31:50.605: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:31:52.700: INFO @metrics_hook: valid matching accuracy: 0.88864903759669, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:31:53.136: INFO @decay_lr : LR updated to `0.000437329` 2019-10-09 09:31:53.137: INFO @log_profile : T train: 180.353485 2019-10-09 09:31:53.138: INFO @log_profile : T valid: 8.509543 2019-10-09 09:31:53.138: INFO @log_profile : T read data: 1.017227 2019-10-09 09:31:53.138: INFO @log_profile : T hooks: 3.137680 2019-10-09 09:31:53.138: INFO @main_loop : Epoch 165 done 2019-10-09 09:31:53.138: INFO @main_loop : Training epoch 166 2019-10-09 09:35:03.385: INFO @log_variables: train loss mean: 0.251312 2019-10-09 09:35:03.385: INFO @log_variables: train age_loss mean: 4.202800 2019-10-09 09:35:03.385: INFO @log_variables: train gender_loss mean: 0.036045 2019-10-09 09:35:03.385: INFO @log_variables: train matching_loss nanmean: 0.322743 2019-10-09 09:35:03.385: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:35:03.385: INFO @log_variables: train age_mae mean: 4.677423 2019-10-09 09:35:03.385: INFO @log_variables: train gender_accuracy mean: 0.987300 2019-10-09 09:35:03.385: INFO @log_variables: train positive_distance nanmean: 0.745319 2019-10-09 09:35:03.385: INFO @log_variables: train negative_distance nanmean: 1.407900 2019-10-09 09:35:03.386: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:35:03.386: INFO @log_variables: valid loss mean: 0.456175 2019-10-09 09:35:03.386: INFO @log_variables: valid age_loss mean: 6.433707 2019-10-09 09:35:03.386: INFO @log_variables: valid gender_loss mean: 0.275680 2019-10-09 09:35:03.386: INFO @log_variables: valid matching_loss nanmean: 0.495091 2019-10-09 09:35:03.386: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:35:03.386: INFO @log_variables: valid age_mae mean: 6.916474 2019-10-09 09:35:03.386: INFO @log_variables: valid gender_accuracy mean: 0.924769 2019-10-09 09:35:03.386: INFO @log_variables: valid positive_distance nanmean: 0.786296 2019-10-09 09:35:03.386: INFO @log_variables: valid negative_distance nanmean: 1.375449 2019-10-09 09:35:03.386: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:35:03.386: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:35:05.668: INFO @metrics_hook: valid matching accuracy: 0.8884091863044912, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:35:06.123: INFO @decay_lr : LR updated to `0.00043514234` 2019-10-09 09:35:06.124: INFO @log_profile : T train: 179.625749 2019-10-09 09:35:06.124: INFO @log_profile : T valid: 8.525799 2019-10-09 09:35:06.124: INFO @log_profile : T read data: 1.435993 2019-10-09 09:35:06.125: INFO @log_profile : T hooks: 3.313704 2019-10-09 09:35:06.125: INFO @main_loop : Epoch 166 done 2019-10-09 09:35:06.125: INFO @main_loop : Training epoch 167 2019-10-09 09:38:16.016: INFO @log_variables: train loss mean: 0.250305 2019-10-09 09:38:16.016: INFO @log_variables: train age_loss mean: 4.162217 2019-10-09 09:38:16.016: INFO @log_variables: train gender_loss mean: 0.039338 2019-10-09 09:38:16.016: INFO @log_variables: train matching_loss nanmean: 0.320385 2019-10-09 09:38:16.016: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:38:16.016: INFO @log_variables: train age_mae mean: 4.636953 2019-10-09 09:38:16.016: INFO @log_variables: train gender_accuracy mean: 0.986152 2019-10-09 09:38:16.016: INFO @log_variables: train positive_distance nanmean: 0.745070 2019-10-09 09:38:16.016: INFO @log_variables: train negative_distance nanmean: 1.407921 2019-10-09 09:38:16.016: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:38:16.017: INFO @log_variables: valid loss mean: 0.456499 2019-10-09 09:38:16.017: INFO @log_variables: valid age_loss mean: 6.343872 2019-10-09 09:38:16.017: INFO @log_variables: valid gender_loss mean: 0.279597 2019-10-09 09:38:16.017: INFO @log_variables: valid matching_loss nanmean: 0.501161 2019-10-09 09:38:16.017: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:38:16.017: INFO @log_variables: valid age_mae mean: 6.825246 2019-10-09 09:38:16.017: INFO @log_variables: valid gender_accuracy mean: 0.923054 2019-10-09 09:38:16.017: INFO @log_variables: valid positive_distance nanmean: 0.783890 2019-10-09 09:38:16.017: INFO @log_variables: valid negative_distance nanmean: 1.373230 2019-10-09 09:38:16.017: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:38:16.017: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:38:18.274: INFO @metrics_hook: valid matching accuracy: 0.8891887030041374, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:38:18.710: INFO @decay_lr : LR updated to `0.00043296663` 2019-10-09 09:38:18.712: INFO @log_profile : T train: 179.592720 2019-10-09 09:38:18.712: INFO @log_profile : T valid: 8.609624 2019-10-09 09:38:18.712: INFO @log_profile : T read data: 1.028234 2019-10-09 09:38:18.712: INFO @log_profile : T hooks: 3.269757 2019-10-09 09:38:18.712: INFO @main_loop : Epoch 167 done 2019-10-09 09:38:18.712: INFO @main_loop : Training epoch 168 2019-10-09 09:41:29.173: INFO @log_variables: train loss mean: 0.249337 2019-10-09 09:41:29.173: INFO @log_variables: train age_loss mean: 4.154129 2019-10-09 09:41:29.173: INFO @log_variables: train gender_loss mean: 0.036968 2019-10-09 09:41:29.173: INFO @log_variables: train matching_loss nanmean: 0.320563 2019-10-09 09:41:29.173: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:41:29.173: INFO @log_variables: train age_mae mean: 4.628766 2019-10-09 09:41:29.173: INFO @log_variables: train gender_accuracy mean: 0.987065 2019-10-09 09:41:29.173: INFO @log_variables: train positive_distance nanmean: 0.744047 2019-10-09 09:41:29.173: INFO @log_variables: train negative_distance nanmean: 1.408163 2019-10-09 09:41:29.173: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:41:29.173: INFO @log_variables: valid loss mean: 0.460234 2019-10-09 09:41:29.173: INFO @log_variables: valid age_loss mean: 6.338313 2019-10-09 09:41:29.174: INFO @log_variables: valid gender_loss mean: 0.291286 2019-10-09 09:41:29.174: INFO @log_variables: valid matching_loss nanmean: 0.501608 2019-10-09 09:41:29.174: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:41:29.174: INFO @log_variables: valid age_mae mean: 6.820949 2019-10-09 09:41:29.174: INFO @log_variables: valid gender_accuracy mean: 0.920334 2019-10-09 09:41:29.174: INFO @log_variables: valid positive_distance nanmean: 0.787078 2019-10-09 09:41:29.174: INFO @log_variables: valid negative_distance nanmean: 1.375573 2019-10-09 09:41:29.174: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:41:29.174: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:41:31.517: INFO @metrics_hook: valid matching accuracy: 0.8889488517119386, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:41:31.946: INFO @decay_lr : LR updated to `0.0004308018` 2019-10-09 09:41:31.947: INFO @log_profile : T train: 179.786728 2019-10-09 09:41:31.947: INFO @log_profile : T valid: 8.557899 2019-10-09 09:41:31.947: INFO @log_profile : T read data: 1.447538 2019-10-09 09:41:31.947: INFO @log_profile : T hooks: 3.358281 2019-10-09 09:41:31.947: INFO @main_loop : Epoch 168 done 2019-10-09 09:41:31.947: INFO @main_loop : Training epoch 169 2019-10-09 09:44:42.326: INFO @log_variables: train loss mean: 0.248337 2019-10-09 09:44:42.327: INFO @log_variables: train age_loss mean: 4.172739 2019-10-09 09:44:42.327: INFO @log_variables: train gender_loss mean: 0.034832 2019-10-09 09:44:42.327: INFO @log_variables: train matching_loss nanmean: 0.317740 2019-10-09 09:44:42.327: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:44:42.327: INFO @log_variables: train age_mae mean: 4.647375 2019-10-09 09:44:42.327: INFO @log_variables: train gender_accuracy mean: 0.987570 2019-10-09 09:44:42.327: INFO @log_variables: train positive_distance nanmean: 0.743838 2019-10-09 09:44:42.327: INFO @log_variables: train negative_distance nanmean: 1.408247 2019-10-09 09:44:42.327: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:44:42.327: INFO @log_variables: valid loss mean: 0.457258 2019-10-09 09:44:42.327: INFO @log_variables: valid age_loss mean: 6.436810 2019-10-09 09:44:42.327: INFO @log_variables: valid gender_loss mean: 0.276382 2019-10-09 09:44:42.327: INFO @log_variables: valid matching_loss nanmean: 0.497438 2019-10-09 09:44:42.327: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:44:42.327: INFO @log_variables: valid age_mae mean: 6.919184 2019-10-09 09:44:42.328: INFO @log_variables: valid gender_accuracy mean: 0.925302 2019-10-09 09:44:42.328: INFO @log_variables: valid positive_distance nanmean: 0.791440 2019-10-09 09:44:42.328: INFO @log_variables: valid negative_distance nanmean: 1.378759 2019-10-09 09:44:42.328: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:44:42.328: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:44:44.612: INFO @metrics_hook: valid matching accuracy: 0.8900881453498831, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:44:45.068: INFO @decay_lr : LR updated to `0.0004286478` 2019-10-09 09:44:45.069: INFO @log_profile : T train: 179.806452 2019-10-09 09:44:45.069: INFO @log_profile : T valid: 8.514489 2019-10-09 09:44:45.069: INFO @log_profile : T read data: 1.392461 2019-10-09 09:44:45.069: INFO @log_profile : T hooks: 3.322549 2019-10-09 09:44:45.069: INFO @main_loop : Epoch 169 done 2019-10-09 09:44:45.069: INFO @main_loop : Training epoch 170 2019-10-09 09:47:55.127: INFO @log_variables: train loss mean: 0.249734 2019-10-09 09:47:55.127: INFO @log_variables: train age_loss mean: 4.136941 2019-10-09 09:47:55.127: INFO @log_variables: train gender_loss mean: 0.038258 2019-10-09 09:47:55.127: INFO @log_variables: train matching_loss nanmean: 0.322224 2019-10-09 09:47:55.127: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:47:55.127: INFO @log_variables: train age_mae mean: 4.611938 2019-10-09 09:47:55.127: INFO @log_variables: train gender_accuracy mean: 0.986620 2019-10-09 09:47:55.127: INFO @log_variables: train positive_distance nanmean: 0.744926 2019-10-09 09:47:55.127: INFO @log_variables: train negative_distance nanmean: 1.408106 2019-10-09 09:47:55.127: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:47:55.128: INFO @log_variables: valid loss mean: 0.457536 2019-10-09 09:47:55.128: INFO @log_variables: valid age_loss mean: 6.419522 2019-10-09 09:47:55.128: INFO @log_variables: valid gender_loss mean: 0.274032 2019-10-09 09:47:55.128: INFO @log_variables: valid matching_loss nanmean: 0.502378 2019-10-09 09:47:55.128: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:47:55.128: INFO @log_variables: valid age_mae mean: 6.903825 2019-10-09 09:47:55.128: INFO @log_variables: valid gender_accuracy mean: 0.922285 2019-10-09 09:47:55.128: INFO @log_variables: valid positive_distance nanmean: 0.785142 2019-10-09 09:47:55.128: INFO @log_variables: valid negative_distance nanmean: 1.374185 2019-10-09 09:47:55.128: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:47:55.128: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:47:57.629: INFO @metrics_hook: valid matching accuracy: 0.8865503387899503, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:47:58.071: INFO @decay_lr : LR updated to `0.00042650456` 2019-10-09 09:47:58.072: INFO @log_profile : T train: 179.865238 2019-10-09 09:47:58.072: INFO @log_profile : T valid: 8.534066 2019-10-09 09:47:58.072: INFO @log_profile : T read data: 0.979950 2019-10-09 09:47:58.072: INFO @log_profile : T hooks: 3.537734 2019-10-09 09:47:58.072: INFO @main_loop : Epoch 170 done 2019-10-09 09:47:58.073: INFO @main_loop : Training epoch 171 2019-10-09 09:51:08.437: INFO @log_variables: train loss mean: 0.247836 2019-10-09 09:51:08.437: INFO @log_variables: train age_loss mean: 4.154902 2019-10-09 09:51:08.437: INFO @log_variables: train gender_loss mean: 0.033973 2019-10-09 09:51:08.437: INFO @log_variables: train matching_loss nanmean: 0.318828 2019-10-09 09:51:08.437: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:51:08.437: INFO @log_variables: train age_mae mean: 4.629782 2019-10-09 09:51:08.437: INFO @log_variables: train gender_accuracy mean: 0.988304 2019-10-09 09:51:08.437: INFO @log_variables: train positive_distance nanmean: 0.744072 2019-10-09 09:51:08.437: INFO @log_variables: train negative_distance nanmean: 1.408285 2019-10-09 09:51:08.437: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:51:08.438: INFO @log_variables: valid loss mean: 0.452750 2019-10-09 09:51:08.438: INFO @log_variables: valid age_loss mean: 6.348470 2019-10-09 09:51:08.438: INFO @log_variables: valid gender_loss mean: 0.269597 2019-10-09 09:51:08.438: INFO @log_variables: valid matching_loss nanmean: 0.499081 2019-10-09 09:51:08.438: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:51:08.438: INFO @log_variables: valid age_mae mean: 6.830017 2019-10-09 09:51:08.438: INFO @log_variables: valid gender_accuracy mean: 0.924178 2019-10-09 09:51:08.438: INFO @log_variables: valid positive_distance nanmean: 0.789446 2019-10-09 09:51:08.438: INFO @log_variables: valid negative_distance nanmean: 1.376862 2019-10-09 09:51:08.438: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:51:08.438: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:51:10.696: INFO @metrics_hook: valid matching accuracy: 0.8898482940576843, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:51:11.124: INFO @decay_lr : LR updated to `0.00042437203` 2019-10-09 09:51:11.125: INFO @log_profile : T train: 179.771721 2019-10-09 09:51:11.125: INFO @log_profile : T valid: 8.512474 2019-10-09 09:51:11.126: INFO @log_profile : T read data: 1.431977 2019-10-09 09:51:11.126: INFO @log_profile : T hooks: 3.250546 2019-10-09 09:51:11.126: INFO @main_loop : Epoch 171 done 2019-10-09 09:51:11.126: INFO @main_loop : Training epoch 172 2019-10-09 09:54:21.403: INFO @log_variables: train loss mean: 0.249153 2019-10-09 09:54:21.403: INFO @log_variables: train age_loss mean: 4.148660 2019-10-09 09:54:21.403: INFO @log_variables: train gender_loss mean: 0.037064 2019-10-09 09:54:21.403: INFO @log_variables: train matching_loss nanmean: 0.320444 2019-10-09 09:54:21.403: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:54:21.403: INFO @log_variables: train age_mae mean: 4.623435 2019-10-09 09:54:21.403: INFO @log_variables: train gender_accuracy mean: 0.986749 2019-10-09 09:54:21.403: INFO @log_variables: train positive_distance nanmean: 0.743609 2019-10-09 09:54:21.403: INFO @log_variables: train negative_distance nanmean: 1.408393 2019-10-09 09:54:21.404: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:54:21.404: INFO @log_variables: valid loss mean: 0.456152 2019-10-09 09:54:21.404: INFO @log_variables: valid age_loss mean: 6.560676 2019-10-09 09:54:21.404: INFO @log_variables: valid gender_loss mean: 0.255283 2019-10-09 09:54:21.404: INFO @log_variables: valid matching_loss nanmean: 0.502720 2019-10-09 09:54:21.404: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:54:21.404: INFO @log_variables: valid age_mae mean: 7.044257 2019-10-09 09:54:21.404: INFO @log_variables: valid gender_accuracy mean: 0.928850 2019-10-09 09:54:21.404: INFO @log_variables: valid positive_distance nanmean: 0.780717 2019-10-09 09:54:21.404: INFO @log_variables: valid negative_distance nanmean: 1.373437 2019-10-09 09:54:21.404: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:54:21.404: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:54:23.932: INFO @metrics_hook: valid matching accuracy: 0.8881093721892427, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:54:24.380: INFO @decay_lr : LR updated to `0.00042225016` 2019-10-09 09:54:25.111: INFO @model : Quantizing and saving the model 2019-10-09 09:54:26.327: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:54:26.333: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:54:26.339: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:54:26.344: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:54:26.350: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:54:26.355: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:54:26.360: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:54:26.365: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:54:26.370: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:54:26.376: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:54:26.382: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:54:26.387: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:54:26.392: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:54:26.397: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:54:26.403: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:54:26.408: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:54:26.413: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:54:26.418: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:54:26.423: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:54:26.429: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:54:26.434: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:54:26.439: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:54:26.445: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:54:26.450: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:54:26.455: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:54:26.460: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:54:26.465: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 09:54:26.474: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 09:54:41.250: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 09:54:41.569: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 09:54:41.591: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 09:54:43.648: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 09:54:43.694: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 09:54:43.699: INFO @log_profile : T train: 179.686637 2019-10-09 09:54:43.702: INFO @log_profile : T valid: 8.538141 2019-10-09 09:54:43.702: INFO @log_profile : T read data: 1.383105 2019-10-09 09:54:43.703: INFO @log_profile : T hooks: 22.880574 2019-10-09 09:54:43.703: INFO @main_loop : Epoch 172 done 2019-10-09 09:54:43.703: INFO @main_loop : Training epoch 173 2019-10-09 09:57:53.615: INFO @log_variables: train loss mean: 0.248636 2019-10-09 09:57:53.615: INFO @log_variables: train age_loss mean: 4.155901 2019-10-09 09:57:53.615: INFO @log_variables: train gender_loss mean: 0.037179 2019-10-09 09:57:53.615: INFO @log_variables: train matching_loss nanmean: 0.318002 2019-10-09 09:57:53.615: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 09:57:53.615: INFO @log_variables: train age_mae mean: 4.630424 2019-10-09 09:57:53.615: INFO @log_variables: train gender_accuracy mean: 0.986883 2019-10-09 09:57:53.615: INFO @log_variables: train positive_distance nanmean: 0.742536 2019-10-09 09:57:53.615: INFO @log_variables: train negative_distance nanmean: 1.408401 2019-10-09 09:57:53.615: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 09:57:53.616: INFO @log_variables: valid loss mean: 0.444070 2019-10-09 09:57:53.616: INFO @log_variables: valid age_loss mean: 6.328136 2019-10-09 09:57:53.616: INFO @log_variables: valid gender_loss mean: 0.241918 2019-10-09 09:57:53.616: INFO @log_variables: valid matching_loss nanmean: 0.501886 2019-10-09 09:57:53.616: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 09:57:53.616: INFO @log_variables: valid age_mae mean: 6.810809 2019-10-09 09:57:53.616: INFO @log_variables: valid gender_accuracy mean: 0.928969 2019-10-09 09:57:53.616: INFO @log_variables: valid positive_distance nanmean: 0.782475 2019-10-09 09:57:53.616: INFO @log_variables: valid negative_distance nanmean: 1.374441 2019-10-09 09:57:53.616: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 09:57:53.616: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 09:57:55.865: INFO @metrics_hook: valid matching accuracy: 0.8881093721892427, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 09:57:56.327: INFO @decay_lr : LR updated to `0.0004201389` 2019-10-09 09:57:57.080: INFO @model : Quantizing and saving the model 2019-10-09 09:57:57.965: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:57:57.971: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:57:57.976: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:57:57.983: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:57:57.988: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:57:57.994: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:57:57.999: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:57:58.005: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:57:58.010: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:57:58.016: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:57:58.022: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:57:58.028: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:57:58.033: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:57:58.039: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:57:58.044: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:57:58.050: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:57:58.056: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:57:58.061: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:57:58.067: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:57:58.073: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:57:58.078: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:57:58.084: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:57:58.089: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:57:58.095: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:57:58.100: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 09:57:58.106: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 09:57:58.112: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 09:57:58.120: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 09:58:11.306: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 09:58:11.610: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 09:58:11.631: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 09:58:13.743: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 09:58:13.790: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 09:58:13.795: INFO @log_profile : T train: 179.654467 2019-10-09 09:58:13.795: INFO @log_profile : T valid: 8.558516 2019-10-09 09:58:13.795: INFO @log_profile : T read data: 1.030390 2019-10-09 09:58:13.795: INFO @log_profile : T hooks: 20.761022 2019-10-09 09:58:13.795: INFO @main_loop : Epoch 173 done 2019-10-09 09:58:13.795: INFO @main_loop : Training epoch 174 2019-10-09 10:01:24.146: INFO @log_variables: train loss mean: 0.248387 2019-10-09 10:01:24.146: INFO @log_variables: train age_loss mean: 4.170732 2019-10-09 10:01:24.146: INFO @log_variables: train gender_loss mean: 0.035237 2019-10-09 10:01:24.146: INFO @log_variables: train matching_loss nanmean: 0.317690 2019-10-09 10:01:24.146: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:01:24.146: INFO @log_variables: train age_mae mean: 4.645255 2019-10-09 10:01:24.146: INFO @log_variables: train gender_accuracy mean: 0.987465 2019-10-09 10:01:24.146: INFO @log_variables: train positive_distance nanmean: 0.742516 2019-10-09 10:01:24.147: INFO @log_variables: train negative_distance nanmean: 1.408212 2019-10-09 10:01:24.147: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:01:24.147: INFO @log_variables: valid loss mean: 0.454580 2019-10-09 10:01:24.147: INFO @log_variables: valid age_loss mean: 6.486389 2019-10-09 10:01:24.147: INFO @log_variables: valid gender_loss mean: 0.261902 2019-10-09 10:01:24.147: INFO @log_variables: valid matching_loss nanmean: 0.498658 2019-10-09 10:01:24.147: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:01:24.147: INFO @log_variables: valid age_mae mean: 6.969489 2019-10-09 10:01:24.147: INFO @log_variables: valid gender_accuracy mean: 0.926307 2019-10-09 10:01:24.147: INFO @log_variables: valid positive_distance nanmean: 0.784468 2019-10-09 10:01:24.147: INFO @log_variables: valid negative_distance nanmean: 1.374977 2019-10-09 10:01:24.147: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:01:24.147: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:01:26.465: INFO @metrics_hook: valid matching accuracy: 0.8891887030041374, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:01:26.919: INFO @decay_lr : LR updated to `0.00041803822` 2019-10-09 10:01:26.920: INFO @log_profile : T train: 179.636106 2019-10-09 10:01:26.920: INFO @log_profile : T valid: 8.599956 2019-10-09 10:01:26.920: INFO @log_profile : T read data: 1.459254 2019-10-09 10:01:26.920: INFO @log_profile : T hooks: 3.344083 2019-10-09 10:01:26.920: INFO @main_loop : Epoch 174 done 2019-10-09 10:01:26.920: INFO @main_loop : Training epoch 175 2019-10-09 10:04:36.528: INFO @log_variables: train loss mean: 0.246526 2019-10-09 10:04:36.528: INFO @log_variables: train age_loss mean: 4.112890 2019-10-09 10:04:36.528: INFO @log_variables: train gender_loss mean: 0.036831 2019-10-09 10:04:36.528: INFO @log_variables: train matching_loss nanmean: 0.316109 2019-10-09 10:04:36.529: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:04:36.529: INFO @log_variables: train age_mae mean: 4.587170 2019-10-09 10:04:36.529: INFO @log_variables: train gender_accuracy mean: 0.987013 2019-10-09 10:04:36.529: INFO @log_variables: train positive_distance nanmean: 0.740971 2019-10-09 10:04:36.529: INFO @log_variables: train negative_distance nanmean: 1.408014 2019-10-09 10:04:36.529: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:04:36.529: INFO @log_variables: valid loss mean: 0.449106 2019-10-09 10:04:36.529: INFO @log_variables: valid age_loss mean: 6.353881 2019-10-09 10:04:36.529: INFO @log_variables: valid gender_loss mean: 0.260412 2019-10-09 10:04:36.529: INFO @log_variables: valid matching_loss nanmean: 0.496427 2019-10-09 10:04:36.529: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:04:36.529: INFO @log_variables: valid age_mae mean: 6.836068 2019-10-09 10:04:36.529: INFO @log_variables: valid gender_accuracy mean: 0.928436 2019-10-09 10:04:36.529: INFO @log_variables: valid positive_distance nanmean: 0.787624 2019-10-09 10:04:36.529: INFO @log_variables: valid negative_distance nanmean: 1.378309 2019-10-09 10:04:36.529: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:04:36.529: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:04:38.690: INFO @metrics_hook: valid matching accuracy: 0.8888888888888888, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:04:39.120: INFO @decay_lr : LR updated to `0.00041594804` 2019-10-09 10:04:39.122: INFO @log_profile : T train: 179.470993 2019-10-09 10:04:39.122: INFO @log_profile : T valid: 8.502024 2019-10-09 10:04:39.122: INFO @log_profile : T read data: 0.992389 2019-10-09 10:04:39.122: INFO @log_profile : T hooks: 3.151368 2019-10-09 10:04:39.122: INFO @main_loop : Epoch 175 done 2019-10-09 10:04:39.122: INFO @main_loop : Training epoch 176 2019-10-09 10:07:49.586: INFO @log_variables: train loss mean: 0.248170 2019-10-09 10:07:49.586: INFO @log_variables: train age_loss mean: 4.139540 2019-10-09 10:07:49.586: INFO @log_variables: train gender_loss mean: 0.036623 2019-10-09 10:07:49.586: INFO @log_variables: train matching_loss nanmean: 0.318750 2019-10-09 10:07:49.586: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:07:49.586: INFO @log_variables: train age_mae mean: 4.613837 2019-10-09 10:07:49.586: INFO @log_variables: train gender_accuracy mean: 0.986964 2019-10-09 10:07:49.586: INFO @log_variables: train positive_distance nanmean: 0.744401 2019-10-09 10:07:49.587: INFO @log_variables: train negative_distance nanmean: 1.408134 2019-10-09 10:07:49.587: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:07:49.587: INFO @log_variables: valid loss mean: 0.457818 2019-10-09 10:07:49.587: INFO @log_variables: valid age_loss mean: 6.449198 2019-10-09 10:07:49.587: INFO @log_variables: valid gender_loss mean: 0.276071 2019-10-09 10:07:49.587: INFO @log_variables: valid matching_loss nanmean: 0.498243 2019-10-09 10:07:49.587: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:07:49.587: INFO @log_variables: valid age_mae mean: 6.932871 2019-10-09 10:07:49.587: INFO @log_variables: valid gender_accuracy mean: 0.924237 2019-10-09 10:07:49.587: INFO @log_variables: valid positive_distance nanmean: 0.784397 2019-10-09 10:07:49.587: INFO @log_variables: valid negative_distance nanmean: 1.375551 2019-10-09 10:07:49.587: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:07:49.587: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:07:52.133: INFO @metrics_hook: valid matching accuracy: 0.8891287401810877, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:07:52.578: INFO @decay_lr : LR updated to `0.0004138683` 2019-10-09 10:07:52.580: INFO @log_profile : T train: 179.872647 2019-10-09 10:07:52.580: INFO @log_profile : T valid: 8.513909 2019-10-09 10:07:52.580: INFO @log_profile : T read data: 1.453963 2019-10-09 10:07:52.580: INFO @log_profile : T hooks: 3.533265 2019-10-09 10:07:52.580: INFO @main_loop : Epoch 176 done 2019-10-09 10:07:52.580: INFO @main_loop : Training epoch 177 2019-10-09 10:11:03.092: INFO @log_variables: train loss mean: 0.244673 2019-10-09 10:11:03.092: INFO @log_variables: train age_loss mean: 4.098703 2019-10-09 10:11:03.092: INFO @log_variables: train gender_loss mean: 0.034321 2019-10-09 10:11:03.092: INFO @log_variables: train matching_loss nanmean: 0.314293 2019-10-09 10:11:03.092: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:11:03.093: INFO @log_variables: train age_mae mean: 4.573439 2019-10-09 10:11:03.093: INFO @log_variables: train gender_accuracy mean: 0.988239 2019-10-09 10:11:03.093: INFO @log_variables: train positive_distance nanmean: 0.740938 2019-10-09 10:11:03.093: INFO @log_variables: train negative_distance nanmean: 1.408059 2019-10-09 10:11:03.093: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:11:03.093: INFO @log_variables: valid loss mean: 0.452610 2019-10-09 10:11:03.093: INFO @log_variables: valid age_loss mean: 6.371005 2019-10-09 10:11:03.093: INFO @log_variables: valid gender_loss mean: 0.267715 2019-10-09 10:11:03.093: INFO @log_variables: valid matching_loss nanmean: 0.498274 2019-10-09 10:11:03.093: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:11:03.093: INFO @log_variables: valid age_mae mean: 6.853827 2019-10-09 10:11:03.093: INFO @log_variables: valid gender_accuracy mean: 0.923527 2019-10-09 10:11:03.093: INFO @log_variables: valid positive_distance nanmean: 0.791518 2019-10-09 10:11:03.093: INFO @log_variables: valid negative_distance nanmean: 1.377605 2019-10-09 10:11:03.093: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:11:03.093: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:11:05.595: INFO @metrics_hook: valid matching accuracy: 0.8871499670204473, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:11:06.054: INFO @decay_lr : LR updated to `0.00041179897` 2019-10-09 10:11:06.055: INFO @log_profile : T train: 179.898254 2019-10-09 10:11:06.055: INFO @log_profile : T valid: 8.556672 2019-10-09 10:11:06.055: INFO @log_profile : T read data: 1.400446 2019-10-09 10:11:06.056: INFO @log_profile : T hooks: 3.533611 2019-10-09 10:11:06.056: INFO @main_loop : Epoch 177 done 2019-10-09 10:11:06.056: INFO @main_loop : Training epoch 178 2019-10-09 10:14:16.192: INFO @log_variables: train loss mean: 0.246527 2019-10-09 10:14:16.192: INFO @log_variables: train age_loss mean: 4.102296 2019-10-09 10:14:16.192: INFO @log_variables: train gender_loss mean: 0.036507 2019-10-09 10:14:16.192: INFO @log_variables: train matching_loss nanmean: 0.317497 2019-10-09 10:14:16.192: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:14:16.192: INFO @log_variables: train age_mae mean: 4.576821 2019-10-09 10:14:16.192: INFO @log_variables: train gender_accuracy mean: 0.987301 2019-10-09 10:14:16.192: INFO @log_variables: train positive_distance nanmean: 0.743403 2019-10-09 10:14:16.192: INFO @log_variables: train negative_distance nanmean: 1.408089 2019-10-09 10:14:16.192: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:14:16.192: INFO @log_variables: valid loss mean: 0.454099 2019-10-09 10:14:16.192: INFO @log_variables: valid age_loss mean: 6.413950 2019-10-09 10:14:16.192: INFO @log_variables: valid gender_loss mean: 0.271976 2019-10-09 10:14:16.192: INFO @log_variables: valid matching_loss nanmean: 0.494335 2019-10-09 10:14:16.193: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:14:16.193: INFO @log_variables: valid age_mae mean: 6.896508 2019-10-09 10:14:16.193: INFO @log_variables: valid gender_accuracy mean: 0.924710 2019-10-09 10:14:16.193: INFO @log_variables: valid positive_distance nanmean: 0.782975 2019-10-09 10:14:16.193: INFO @log_variables: valid negative_distance nanmean: 1.375577 2019-10-09 10:14:16.193: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:14:16.193: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:14:18.467: INFO @metrics_hook: valid matching accuracy: 0.8890088145349883, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:14:18.920: INFO @decay_lr : LR updated to `0.00040973997` 2019-10-09 10:14:18.921: INFO @log_profile : T train: 179.942295 2019-10-09 10:14:18.921: INFO @log_profile : T valid: 8.538814 2019-10-09 10:14:18.921: INFO @log_profile : T read data: 1.013424 2019-10-09 10:14:18.921: INFO @log_profile : T hooks: 3.285360 2019-10-09 10:14:18.921: INFO @main_loop : Epoch 178 done 2019-10-09 10:14:18.921: INFO @main_loop : Training epoch 179 2019-10-09 10:17:29.092: INFO @log_variables: train loss mean: 0.245933 2019-10-09 10:17:29.093: INFO @log_variables: train age_loss mean: 4.136786 2019-10-09 10:17:29.093: INFO @log_variables: train gender_loss mean: 0.033186 2019-10-09 10:17:29.093: INFO @log_variables: train matching_loss nanmean: 0.315528 2019-10-09 10:17:29.093: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:17:29.093: INFO @log_variables: train age_mae mean: 4.611327 2019-10-09 10:17:29.093: INFO @log_variables: train gender_accuracy mean: 0.988539 2019-10-09 10:17:29.093: INFO @log_variables: train positive_distance nanmean: 0.741607 2019-10-09 10:17:29.093: INFO @log_variables: train negative_distance nanmean: 1.407862 2019-10-09 10:17:29.093: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:17:29.093: INFO @log_variables: valid loss mean: 0.451126 2019-10-09 10:17:29.093: INFO @log_variables: valid age_loss mean: 6.273786 2019-10-09 10:17:29.093: INFO @log_variables: valid gender_loss mean: 0.274209 2019-10-09 10:17:29.093: INFO @log_variables: valid matching_loss nanmean: 0.496904 2019-10-09 10:17:29.093: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:17:29.093: INFO @log_variables: valid age_mae mean: 6.756939 2019-10-09 10:17:29.093: INFO @log_variables: valid gender_accuracy mean: 0.925361 2019-10-09 10:17:29.094: INFO @log_variables: valid positive_distance nanmean: 0.783879 2019-10-09 10:17:29.094: INFO @log_variables: valid negative_distance nanmean: 1.375929 2019-10-09 10:17:29.094: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:17:29.094: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:17:31.559: INFO @metrics_hook: valid matching accuracy: 0.8899682197037837, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:17:32.000: INFO @decay_lr : LR updated to `0.00040769126` 2019-10-09 10:17:32.001: INFO @log_profile : T train: 179.612972 2019-10-09 10:17:32.001: INFO @log_profile : T valid: 8.517739 2019-10-09 10:17:32.001: INFO @log_profile : T read data: 1.379387 2019-10-09 10:17:32.001: INFO @log_profile : T hooks: 3.484790 2019-10-09 10:17:32.001: INFO @main_loop : Epoch 179 done 2019-10-09 10:17:32.002: INFO @main_loop : Training epoch 180 2019-10-09 10:20:42.174: INFO @log_variables: train loss mean: 0.245345 2019-10-09 10:20:42.174: INFO @log_variables: train age_loss mean: 4.092775 2019-10-09 10:20:42.174: INFO @log_variables: train gender_loss mean: 0.036155 2019-10-09 10:20:42.174: INFO @log_variables: train matching_loss nanmean: 0.315138 2019-10-09 10:20:42.174: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:20:42.174: INFO @log_variables: train age_mae mean: 4.566361 2019-10-09 10:20:42.174: INFO @log_variables: train gender_accuracy mean: 0.987093 2019-10-09 10:20:42.175: INFO @log_variables: train positive_distance nanmean: 0.741299 2019-10-09 10:20:42.175: INFO @log_variables: train negative_distance nanmean: 1.407845 2019-10-09 10:20:42.175: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:20:42.175: INFO @log_variables: valid loss mean: 0.449139 2019-10-09 10:20:42.175: INFO @log_variables: valid age_loss mean: 6.435151 2019-10-09 10:20:42.175: INFO @log_variables: valid gender_loss mean: 0.245345 2019-10-09 10:20:42.175: INFO @log_variables: valid matching_loss nanmean: 0.503471 2019-10-09 10:20:42.175: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:20:42.175: INFO @log_variables: valid age_mae mean: 6.918204 2019-10-09 10:20:42.175: INFO @log_variables: valid gender_accuracy mean: 0.928377 2019-10-09 10:20:42.175: INFO @log_variables: valid positive_distance nanmean: 0.782299 2019-10-09 10:20:42.175: INFO @log_variables: valid negative_distance nanmean: 1.373422 2019-10-09 10:20:42.175: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:20:42.175: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:20:44.795: INFO @metrics_hook: valid matching accuracy: 0.8867901900821491, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:20:45.241: INFO @decay_lr : LR updated to `0.0004056528` 2019-10-09 10:20:45.243: INFO @log_profile : T train: 179.579864 2019-10-09 10:20:45.243: INFO @log_profile : T valid: 8.532675 2019-10-09 10:20:45.243: INFO @log_profile : T read data: 1.402708 2019-10-09 10:20:45.243: INFO @log_profile : T hooks: 3.641111 2019-10-09 10:20:45.243: INFO @main_loop : Epoch 180 done 2019-10-09 10:20:45.243: INFO @main_loop : Training epoch 181 2019-10-09 10:23:55.022: INFO @log_variables: train loss mean: 0.246110 2019-10-09 10:23:55.022: INFO @log_variables: train age_loss mean: 4.124137 2019-10-09 10:23:55.022: INFO @log_variables: train gender_loss mean: 0.034403 2019-10-09 10:23:55.022: INFO @log_variables: train matching_loss nanmean: 0.316123 2019-10-09 10:23:55.022: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:23:55.022: INFO @log_variables: train age_mae mean: 4.598342 2019-10-09 10:23:55.022: INFO @log_variables: train gender_accuracy mean: 0.988023 2019-10-09 10:23:55.022: INFO @log_variables: train positive_distance nanmean: 0.742343 2019-10-09 10:23:55.022: INFO @log_variables: train negative_distance nanmean: 1.407994 2019-10-09 10:23:55.022: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:23:55.023: INFO @log_variables: valid loss mean: 0.456295 2019-10-09 10:23:55.023: INFO @log_variables: valid age_loss mean: 6.545388 2019-10-09 10:23:55.023: INFO @log_variables: valid gender_loss mean: 0.261733 2019-10-09 10:23:55.023: INFO @log_variables: valid matching_loss nanmean: 0.498244 2019-10-09 10:23:55.023: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:23:55.023: INFO @log_variables: valid age_mae mean: 7.028771 2019-10-09 10:23:55.023: INFO @log_variables: valid gender_accuracy mean: 0.927490 2019-10-09 10:23:55.023: INFO @log_variables: valid positive_distance nanmean: 0.785109 2019-10-09 10:23:55.023: INFO @log_variables: valid negative_distance nanmean: 1.377304 2019-10-09 10:23:55.023: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:23:55.023: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:23:57.338: INFO @metrics_hook: valid matching accuracy: 0.8897283684115849, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:23:57.785: INFO @decay_lr : LR updated to `0.00040362452` 2019-10-09 10:23:57.786: INFO @log_profile : T train: 179.681752 2019-10-09 10:23:57.786: INFO @log_profile : T valid: 8.477409 2019-10-09 10:23:57.786: INFO @log_profile : T read data: 0.972207 2019-10-09 10:23:57.786: INFO @log_profile : T hooks: 3.327140 2019-10-09 10:23:57.787: INFO @main_loop : Epoch 181 done 2019-10-09 10:23:57.787: INFO @main_loop : Training epoch 182 2019-10-09 10:27:08.284: INFO @log_variables: train loss mean: 0.246841 2019-10-09 10:27:08.284: INFO @log_variables: train age_loss mean: 4.115952 2019-10-09 10:27:08.284: INFO @log_variables: train gender_loss mean: 0.036651 2019-10-09 10:27:08.284: INFO @log_variables: train matching_loss nanmean: 0.316961 2019-10-09 10:27:08.284: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:27:08.284: INFO @log_variables: train age_mae mean: 4.590293 2019-10-09 10:27:08.284: INFO @log_variables: train gender_accuracy mean: 0.987227 2019-10-09 10:27:08.284: INFO @log_variables: train positive_distance nanmean: 0.742325 2019-10-09 10:27:08.285: INFO @log_variables: train negative_distance nanmean: 1.408039 2019-10-09 10:27:08.285: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:27:08.285: INFO @log_variables: valid loss mean: 0.454071 2019-10-09 10:27:08.285: INFO @log_variables: valid age_loss mean: 6.416821 2019-10-09 10:27:08.285: INFO @log_variables: valid gender_loss mean: 0.273215 2019-10-09 10:27:08.285: INFO @log_variables: valid matching_loss nanmean: 0.492725 2019-10-09 10:27:08.285: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:27:08.285: INFO @log_variables: valid age_mae mean: 6.899524 2019-10-09 10:27:08.285: INFO @log_variables: valid gender_accuracy mean: 0.925420 2019-10-09 10:27:08.285: INFO @log_variables: valid positive_distance nanmean: 0.790371 2019-10-09 10:27:08.285: INFO @log_variables: valid negative_distance nanmean: 1.380041 2019-10-09 10:27:08.285: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:27:08.285: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:27:10.764: INFO @metrics_hook: valid matching accuracy: 0.8881693350122923, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:27:11.222: INFO @decay_lr : LR updated to `0.0004016064` 2019-10-09 10:27:11.224: INFO @log_profile : T train: 179.908557 2019-10-09 10:27:11.224: INFO @log_profile : T valid: 8.538404 2019-10-09 10:27:11.224: INFO @log_profile : T read data: 1.403028 2019-10-09 10:27:11.224: INFO @log_profile : T hooks: 3.501429 2019-10-09 10:27:11.224: INFO @main_loop : Epoch 182 done 2019-10-09 10:27:11.224: INFO @main_loop : Training epoch 183 2019-10-09 10:30:21.134: INFO @log_variables: train loss mean: 0.246147 2019-10-09 10:30:21.134: INFO @log_variables: train age_loss mean: 4.119967 2019-10-09 10:30:21.134: INFO @log_variables: train gender_loss mean: 0.033345 2019-10-09 10:30:21.134: INFO @log_variables: train matching_loss nanmean: 0.317715 2019-10-09 10:30:21.134: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:30:21.134: INFO @log_variables: train age_mae mean: 4.594139 2019-10-09 10:30:21.134: INFO @log_variables: train gender_accuracy mean: 0.988702 2019-10-09 10:30:21.134: INFO @log_variables: train positive_distance nanmean: 0.742902 2019-10-09 10:30:21.134: INFO @log_variables: train negative_distance nanmean: 1.407612 2019-10-09 10:30:21.134: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:30:21.134: INFO @log_variables: valid loss mean: 0.449552 2019-10-09 10:30:21.135: INFO @log_variables: valid age_loss mean: 6.390930 2019-10-09 10:30:21.135: INFO @log_variables: valid gender_loss mean: 0.257158 2019-10-09 10:30:21.135: INFO @log_variables: valid matching_loss nanmean: 0.497361 2019-10-09 10:30:21.135: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:30:21.135: INFO @log_variables: valid age_mae mean: 6.874463 2019-10-09 10:30:21.135: INFO @log_variables: valid gender_accuracy mean: 0.928850 2019-10-09 10:30:21.135: INFO @log_variables: valid positive_distance nanmean: 0.784820 2019-10-09 10:30:21.135: INFO @log_variables: valid negative_distance nanmean: 1.376119 2019-10-09 10:30:21.135: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:30:21.135: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:30:23.615: INFO @metrics_hook: valid matching accuracy: 0.8896084427654855, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:30:24.053: INFO @decay_lr : LR updated to `0.0003995984` 2019-10-09 10:30:24.054: INFO @log_profile : T train: 179.787169 2019-10-09 10:30:24.054: INFO @log_profile : T valid: 8.488692 2019-10-09 10:30:24.054: INFO @log_profile : T read data: 0.972234 2019-10-09 10:30:24.054: INFO @log_profile : T hooks: 3.495230 2019-10-09 10:30:24.054: INFO @main_loop : Epoch 183 done 2019-10-09 10:30:24.054: INFO @main_loop : Training epoch 184 2019-10-09 10:33:34.703: INFO @log_variables: train loss mean: 0.245613 2019-10-09 10:33:34.703: INFO @log_variables: train age_loss mean: 4.109089 2019-10-09 10:33:34.703: INFO @log_variables: train gender_loss mean: 0.034088 2019-10-09 10:33:34.703: INFO @log_variables: train matching_loss nanmean: 0.316402 2019-10-09 10:33:34.703: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:33:34.703: INFO @log_variables: train age_mae mean: 4.583553 2019-10-09 10:33:34.703: INFO @log_variables: train gender_accuracy mean: 0.988170 2019-10-09 10:33:34.703: INFO @log_variables: train positive_distance nanmean: 0.742461 2019-10-09 10:33:34.703: INFO @log_variables: train negative_distance nanmean: 1.407955 2019-10-09 10:33:34.704: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:33:34.704: INFO @log_variables: valid loss mean: 0.460195 2019-10-09 10:33:34.704: INFO @log_variables: valid age_loss mean: 6.516155 2019-10-09 10:33:34.704: INFO @log_variables: valid gender_loss mean: 0.275002 2019-10-09 10:33:34.704: INFO @log_variables: valid matching_loss nanmean: 0.499985 2019-10-09 10:33:34.704: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:33:34.704: INFO @log_variables: valid age_mae mean: 6.999871 2019-10-09 10:33:34.704: INFO @log_variables: valid gender_accuracy mean: 0.927372 2019-10-09 10:33:34.704: INFO @log_variables: valid positive_distance nanmean: 0.786691 2019-10-09 10:33:34.704: INFO @log_variables: valid negative_distance nanmean: 1.375376 2019-10-09 10:33:34.704: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:33:34.704: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:33:37.012: INFO @metrics_hook: valid matching accuracy: 0.8900281825268334, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:33:37.454: INFO @decay_lr : LR updated to `0.0003976004` 2019-10-09 10:33:37.455: INFO @log_profile : T train: 179.940212 2019-10-09 10:33:37.456: INFO @log_profile : T valid: 8.560827 2019-10-09 10:33:37.456: INFO @log_profile : T read data: 1.486117 2019-10-09 10:33:37.456: INFO @log_profile : T hooks: 3.327469 2019-10-09 10:33:37.456: INFO @main_loop : Epoch 184 done 2019-10-09 10:33:37.456: INFO @main_loop : Training epoch 185 2019-10-09 10:36:47.678: INFO @log_variables: train loss mean: 0.246195 2019-10-09 10:36:47.678: INFO @log_variables: train age_loss mean: 4.085710 2019-10-09 10:36:47.678: INFO @log_variables: train gender_loss mean: 0.038431 2019-10-09 10:36:47.678: INFO @log_variables: train matching_loss nanmean: 0.316202 2019-10-09 10:36:47.678: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:36:47.678: INFO @log_variables: train age_mae mean: 4.560001 2019-10-09 10:36:47.678: INFO @log_variables: train gender_accuracy mean: 0.986000 2019-10-09 10:36:47.678: INFO @log_variables: train positive_distance nanmean: 0.742262 2019-10-09 10:36:47.678: INFO @log_variables: train negative_distance nanmean: 1.407812 2019-10-09 10:36:47.678: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:36:47.678: INFO @log_variables: valid loss mean: 0.451832 2019-10-09 10:36:47.679: INFO @log_variables: valid age_loss mean: 6.473005 2019-10-09 10:36:47.679: INFO @log_variables: valid gender_loss mean: 0.256658 2019-10-09 10:36:47.679: INFO @log_variables: valid matching_loss nanmean: 0.496722 2019-10-09 10:36:47.679: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:36:47.679: INFO @log_variables: valid age_mae mean: 6.955613 2019-10-09 10:36:47.679: INFO @log_variables: valid gender_accuracy mean: 0.927963 2019-10-09 10:36:47.679: INFO @log_variables: valid positive_distance nanmean: 0.787665 2019-10-09 10:36:47.679: INFO @log_variables: valid negative_distance nanmean: 1.378914 2019-10-09 10:36:47.679: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:36:47.679: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:36:50.220: INFO @metrics_hook: valid matching accuracy: 0.8889488517119386, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:36:50.672: INFO @decay_lr : LR updated to `0.0003956124` 2019-10-09 10:36:50.673: INFO @log_profile : T train: 179.622194 2019-10-09 10:36:50.673: INFO @log_profile : T valid: 8.540989 2019-10-09 10:36:50.673: INFO @log_profile : T read data: 1.415879 2019-10-09 10:36:50.673: INFO @log_profile : T hooks: 3.553434 2019-10-09 10:36:50.673: INFO @main_loop : Epoch 185 done 2019-10-09 10:36:50.673: INFO @main_loop : Training epoch 186 2019-10-09 10:40:00.736: INFO @log_variables: train loss mean: 0.244938 2019-10-09 10:40:00.736: INFO @log_variables: train age_loss mean: 4.094948 2019-10-09 10:40:00.736: INFO @log_variables: train gender_loss mean: 0.034143 2019-10-09 10:40:00.736: INFO @log_variables: train matching_loss nanmean: 0.315669 2019-10-09 10:40:00.736: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:40:00.736: INFO @log_variables: train age_mae mean: 4.569064 2019-10-09 10:40:00.736: INFO @log_variables: train gender_accuracy mean: 0.987951 2019-10-09 10:40:00.736: INFO @log_variables: train positive_distance nanmean: 0.741133 2019-10-09 10:40:00.736: INFO @log_variables: train negative_distance nanmean: 1.407896 2019-10-09 10:40:00.736: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:40:00.736: INFO @log_variables: valid loss mean: 0.447824 2019-10-09 10:40:00.737: INFO @log_variables: valid age_loss mean: 6.337679 2019-10-09 10:40:00.737: INFO @log_variables: valid gender_loss mean: 0.254444 2019-10-09 10:40:00.737: INFO @log_variables: valid matching_loss nanmean: 0.500042 2019-10-09 10:40:00.737: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:40:00.737: INFO @log_variables: valid age_mae mean: 6.819680 2019-10-09 10:40:00.737: INFO @log_variables: valid gender_accuracy mean: 0.929501 2019-10-09 10:40:00.737: INFO @log_variables: valid positive_distance nanmean: 0.776750 2019-10-09 10:40:00.737: INFO @log_variables: valid negative_distance nanmean: 1.372652 2019-10-09 10:40:00.737: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:40:00.737: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:40:02.966: INFO @metrics_hook: valid matching accuracy: 0.8876896324278947, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:40:03.420: INFO @decay_lr : LR updated to `0.00039363434` 2019-10-09 10:40:04.157: INFO @model : Quantizing and saving the model 2019-10-09 10:40:05.369: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:40:05.375: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:40:05.382: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:40:05.388: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:40:05.394: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:40:05.399: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:40:05.407: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:40:05.414: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:40:05.420: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:40:05.427: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:40:05.433: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:40:05.439: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:40:05.444: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:40:05.450: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:40:05.456: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:40:05.461: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:40:05.467: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:40:05.473: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:40:05.479: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:40:05.485: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:40:05.491: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:40:05.497: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:40:05.503: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:40:05.509: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:40:05.517: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:40:05.523: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:40:05.529: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 10:40:05.538: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 10:40:20.833: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 10:40:21.149: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 10:40:21.170: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 10:40:23.266: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 10:40:23.321: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 10:40:23.325: INFO @log_profile : T train: 179.723450 2019-10-09 10:40:23.328: INFO @log_profile : T valid: 8.665408 2019-10-09 10:40:23.329: INFO @log_profile : T read data: 1.007967 2019-10-09 10:40:23.329: INFO @log_profile : T hooks: 23.168712 2019-10-09 10:40:23.329: INFO @main_loop : Epoch 186 done 2019-10-09 10:40:23.329: INFO @main_loop : Training epoch 187 2019-10-09 10:43:33.719: INFO @log_variables: train loss mean: 0.244751 2019-10-09 10:43:33.719: INFO @log_variables: train age_loss mean: 4.085253 2019-10-09 10:43:33.719: INFO @log_variables: train gender_loss mean: 0.034710 2019-10-09 10:43:33.719: INFO @log_variables: train matching_loss nanmean: 0.315494 2019-10-09 10:43:33.719: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:43:33.719: INFO @log_variables: train age_mae mean: 4.559385 2019-10-09 10:43:33.719: INFO @log_variables: train gender_accuracy mean: 0.987832 2019-10-09 10:43:33.719: INFO @log_variables: train positive_distance nanmean: 0.742552 2019-10-09 10:43:33.719: INFO @log_variables: train negative_distance nanmean: 1.408136 2019-10-09 10:43:33.719: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:43:33.720: INFO @log_variables: valid loss mean: 0.452044 2019-10-09 10:43:33.720: INFO @log_variables: valid age_loss mean: 6.319816 2019-10-09 10:43:33.720: INFO @log_variables: valid gender_loss mean: 0.269234 2019-10-09 10:43:33.720: INFO @log_variables: valid matching_loss nanmean: 0.500119 2019-10-09 10:43:33.720: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:43:33.720: INFO @log_variables: valid age_mae mean: 6.802394 2019-10-09 10:43:33.720: INFO @log_variables: valid gender_accuracy mean: 0.926721 2019-10-09 10:43:33.720: INFO @log_variables: valid positive_distance nanmean: 0.784528 2019-10-09 10:43:33.720: INFO @log_variables: valid negative_distance nanmean: 1.377198 2019-10-09 10:43:33.720: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:43:33.720: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:43:35.865: INFO @metrics_hook: valid matching accuracy: 0.8900281825268334, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:43:36.371: INFO @decay_lr : LR updated to `0.00039166617` 2019-10-09 10:43:36.373: INFO @log_profile : T train: 179.756497 2019-10-09 10:43:36.373: INFO @log_profile : T valid: 8.512751 2019-10-09 10:43:36.373: INFO @log_profile : T read data: 1.426262 2019-10-09 10:43:36.373: INFO @log_profile : T hooks: 3.260642 2019-10-09 10:43:36.373: INFO @main_loop : Epoch 187 done 2019-10-09 10:43:36.373: INFO @main_loop : Training epoch 188 2019-10-09 10:46:47.077: INFO @log_variables: train loss mean: 0.242713 2019-10-09 10:46:47.078: INFO @log_variables: train age_loss mean: 4.068409 2019-10-09 10:46:47.078: INFO @log_variables: train gender_loss mean: 0.032615 2019-10-09 10:46:47.078: INFO @log_variables: train matching_loss nanmean: 0.312956 2019-10-09 10:46:47.078: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:46:47.078: INFO @log_variables: train age_mae mean: 4.542500 2019-10-09 10:46:47.078: INFO @log_variables: train gender_accuracy mean: 0.988657 2019-10-09 10:46:47.078: INFO @log_variables: train positive_distance nanmean: 0.740047 2019-10-09 10:46:47.078: INFO @log_variables: train negative_distance nanmean: 1.408304 2019-10-09 10:46:47.078: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:46:47.078: INFO @log_variables: valid loss mean: 0.447901 2019-10-09 10:46:47.078: INFO @log_variables: valid age_loss mean: 6.439267 2019-10-09 10:46:47.078: INFO @log_variables: valid gender_loss mean: 0.248712 2019-10-09 10:46:47.078: INFO @log_variables: valid matching_loss nanmean: 0.495855 2019-10-09 10:46:47.078: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:46:47.078: INFO @log_variables: valid age_mae mean: 6.922524 2019-10-09 10:46:47.078: INFO @log_variables: valid gender_accuracy mean: 0.931689 2019-10-09 10:46:47.078: INFO @log_variables: valid positive_distance nanmean: 0.783616 2019-10-09 10:46:47.078: INFO @log_variables: valid negative_distance nanmean: 1.375478 2019-10-09 10:46:47.079: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:46:47.079: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:46:49.391: INFO @metrics_hook: valid matching accuracy: 0.8894285542963363, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:46:49.866: INFO @decay_lr : LR updated to `0.00038970783` 2019-10-09 10:46:50.631: INFO @model : Quantizing and saving the model 2019-10-09 10:46:51.487: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:46:51.492: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:46:51.498: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:46:51.503: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:46:51.508: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:46:51.513: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:46:51.519: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:46:51.525: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:46:51.530: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:46:51.535: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:46:51.540: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:46:51.546: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:46:51.551: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:46:51.556: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:46:51.561: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:46:51.566: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:46:51.572: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:46:51.577: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:46:51.582: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:46:51.588: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:46:51.595: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:46:51.600: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:46:51.606: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:46:51.611: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:46:51.617: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 10:46:51.623: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 10:46:51.628: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 10:46:51.922: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 10:47:07.448: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 10:47:07.758: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 10:47:07.778: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 10:47:09.847: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 10:47:09.892: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 10:47:09.897: INFO @log_profile : T train: 179.989322 2019-10-09 10:47:09.899: INFO @log_profile : T valid: 8.522315 2019-10-09 10:47:09.902: INFO @log_profile : T read data: 1.516565 2019-10-09 10:47:09.902: INFO @log_profile : T hooks: 23.408480 2019-10-09 10:47:09.902: INFO @main_loop : Epoch 188 done 2019-10-09 10:47:09.902: INFO @main_loop : Training epoch 189 2019-10-09 10:50:20.227: INFO @log_variables: train loss mean: 0.245492 2019-10-09 10:50:20.227: INFO @log_variables: train age_loss mean: 4.091711 2019-10-09 10:50:20.227: INFO @log_variables: train gender_loss mean: 0.037619 2019-10-09 10:50:20.227: INFO @log_variables: train matching_loss nanmean: 0.314235 2019-10-09 10:50:20.227: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:50:20.227: INFO @log_variables: train age_mae mean: 4.565885 2019-10-09 10:50:20.227: INFO @log_variables: train gender_accuracy mean: 0.986937 2019-10-09 10:50:20.227: INFO @log_variables: train positive_distance nanmean: 0.741958 2019-10-09 10:50:20.227: INFO @log_variables: train negative_distance nanmean: 1.408151 2019-10-09 10:50:20.227: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:50:20.227: INFO @log_variables: valid loss mean: 0.445838 2019-10-09 10:50:20.228: INFO @log_variables: valid age_loss mean: 6.375674 2019-10-09 10:50:20.228: INFO @log_variables: valid gender_loss mean: 0.247354 2019-10-09 10:50:20.228: INFO @log_variables: valid matching_loss nanmean: 0.497175 2019-10-09 10:50:20.228: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:50:20.228: INFO @log_variables: valid age_mae mean: 6.859155 2019-10-09 10:50:20.228: INFO @log_variables: valid gender_accuracy mean: 0.929501 2019-10-09 10:50:20.228: INFO @log_variables: valid positive_distance nanmean: 0.786139 2019-10-09 10:50:20.228: INFO @log_variables: valid negative_distance nanmean: 1.377877 2019-10-09 10:50:20.228: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:50:20.228: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:50:22.372: INFO @metrics_hook: valid matching accuracy: 0.8868501529051988, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:50:22.857: INFO @decay_lr : LR updated to `0.0003877593` 2019-10-09 10:50:22.858: INFO @log_profile : T train: 180.029317 2019-10-09 10:50:22.858: INFO @log_profile : T valid: 8.521818 2019-10-09 10:50:22.858: INFO @log_profile : T read data: 1.057550 2019-10-09 10:50:22.859: INFO @log_profile : T hooks: 3.261017 2019-10-09 10:50:22.859: INFO @main_loop : Epoch 189 done 2019-10-09 10:50:22.859: INFO @main_loop : Training epoch 190 2019-10-09 10:53:33.414: INFO @log_variables: train loss mean: 0.243254 2019-10-09 10:53:33.415: INFO @log_variables: train age_loss mean: 4.087279 2019-10-09 10:53:33.415: INFO @log_variables: train gender_loss mean: 0.034059 2019-10-09 10:53:33.415: INFO @log_variables: train matching_loss nanmean: 0.311300 2019-10-09 10:53:33.415: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:53:33.415: INFO @log_variables: train age_mae mean: 4.561817 2019-10-09 10:53:33.415: INFO @log_variables: train gender_accuracy mean: 0.987907 2019-10-09 10:53:33.415: INFO @log_variables: train positive_distance nanmean: 0.739753 2019-10-09 10:53:33.415: INFO @log_variables: train negative_distance nanmean: 1.407913 2019-10-09 10:53:33.415: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:53:33.415: INFO @log_variables: valid loss mean: 0.451235 2019-10-09 10:53:33.415: INFO @log_variables: valid age_loss mean: 6.356877 2019-10-09 10:53:33.415: INFO @log_variables: valid gender_loss mean: 0.267562 2019-10-09 10:53:33.415: INFO @log_variables: valid matching_loss nanmean: 0.495580 2019-10-09 10:53:33.415: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:53:33.415: INFO @log_variables: valid age_mae mean: 6.839995 2019-10-09 10:53:33.415: INFO @log_variables: valid gender_accuracy mean: 0.926662 2019-10-09 10:53:33.415: INFO @log_variables: valid positive_distance nanmean: 0.784943 2019-10-09 10:53:33.415: INFO @log_variables: valid negative_distance nanmean: 1.378639 2019-10-09 10:53:33.416: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:53:33.416: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:53:35.462: INFO @metrics_hook: valid matching accuracy: 0.8881093721892427, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:53:35.956: INFO @decay_lr : LR updated to `0.0003858205` 2019-10-09 10:53:35.957: INFO @log_profile : T train: 179.895409 2019-10-09 10:53:35.957: INFO @log_profile : T valid: 8.515062 2019-10-09 10:53:35.957: INFO @log_profile : T read data: 1.485351 2019-10-09 10:53:35.957: INFO @log_profile : T hooks: 3.117846 2019-10-09 10:53:35.957: INFO @main_loop : Epoch 190 done 2019-10-09 10:53:35.957: INFO @main_loop : Training epoch 191 2019-10-09 10:56:45.415: INFO @log_variables: train loss mean: 0.242630 2019-10-09 10:56:45.415: INFO @log_variables: train age_loss mean: 4.075334 2019-10-09 10:56:45.416: INFO @log_variables: train gender_loss mean: 0.033127 2019-10-09 10:56:45.416: INFO @log_variables: train matching_loss nanmean: 0.311492 2019-10-09 10:56:45.416: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 10:56:45.416: INFO @log_variables: train age_mae mean: 4.549808 2019-10-09 10:56:45.416: INFO @log_variables: train gender_accuracy mean: 0.988640 2019-10-09 10:56:45.416: INFO @log_variables: train positive_distance nanmean: 0.739031 2019-10-09 10:56:45.416: INFO @log_variables: train negative_distance nanmean: 1.407962 2019-10-09 10:56:45.416: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:56:45.416: INFO @log_variables: valid loss mean: 0.447802 2019-10-09 10:56:45.416: INFO @log_variables: valid age_loss mean: 6.310283 2019-10-09 10:56:45.416: INFO @log_variables: valid gender_loss mean: 0.260636 2019-10-09 10:56:45.416: INFO @log_variables: valid matching_loss nanmean: 0.496522 2019-10-09 10:56:45.416: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:56:45.416: INFO @log_variables: valid age_mae mean: 6.793156 2019-10-09 10:56:45.416: INFO @log_variables: valid gender_accuracy mean: 0.930270 2019-10-09 10:56:45.416: INFO @log_variables: valid positive_distance nanmean: 0.786654 2019-10-09 10:56:45.416: INFO @log_variables: valid negative_distance nanmean: 1.377472 2019-10-09 10:56:45.416: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:56:45.416: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 10:56:47.400: INFO @metrics_hook: valid matching accuracy: 0.887629669604845, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 10:56:47.886: INFO @decay_lr : LR updated to `0.0003838914` 2019-10-09 10:56:47.888: INFO @log_profile : T train: 179.107249 2019-10-09 10:56:47.888: INFO @log_profile : T valid: 8.594741 2019-10-09 10:56:47.888: INFO @log_profile : T read data: 1.048437 2019-10-09 10:56:47.888: INFO @log_profile : T hooks: 3.094808 2019-10-09 10:56:47.888: INFO @main_loop : Epoch 191 done 2019-10-09 10:56:47.888: INFO @main_loop : Training epoch 192 2019-10-09 10:59:58.625: INFO @log_variables: train loss mean: 0.241940 2019-10-09 10:59:58.626: INFO @log_variables: train age_loss mean: 4.040994 2019-10-09 10:59:58.626: INFO @log_variables: train gender_loss mean: 0.033408 2019-10-09 10:59:58.626: INFO @log_variables: train matching_loss nanmean: 0.312507 2019-10-09 10:59:58.626: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 10:59:58.626: INFO @log_variables: train age_mae mean: 4.514595 2019-10-09 10:59:58.626: INFO @log_variables: train gender_accuracy mean: 0.988339 2019-10-09 10:59:58.627: INFO @log_variables: train positive_distance nanmean: 0.739989 2019-10-09 10:59:58.627: INFO @log_variables: train negative_distance nanmean: 1.408173 2019-10-09 10:59:58.627: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 10:59:58.627: INFO @log_variables: valid loss mean: 0.460275 2019-10-09 10:59:58.627: INFO @log_variables: valid age_loss mean: 6.336466 2019-10-09 10:59:58.627: INFO @log_variables: valid gender_loss mean: 0.297683 2019-10-09 10:59:58.628: INFO @log_variables: valid matching_loss nanmean: 0.495524 2019-10-09 10:59:58.628: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 10:59:58.628: INFO @log_variables: valid age_mae mean: 6.819324 2019-10-09 10:59:58.628: INFO @log_variables: valid gender_accuracy mean: 0.919742 2019-10-09 10:59:58.628: INFO @log_variables: valid positive_distance nanmean: 0.781463 2019-10-09 10:59:58.628: INFO @log_variables: valid negative_distance nanmean: 1.375820 2019-10-09 10:59:58.629: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 10:59:58.629: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:00:01.694: INFO @metrics_hook: valid matching accuracy: 0.8880494093661929, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:00:02.170: INFO @decay_lr : LR updated to `0.00038197194` 2019-10-09 11:00:02.172: INFO @log_profile : T train: 180.038246 2019-10-09 11:00:02.172: INFO @log_profile : T valid: 8.509209 2019-10-09 11:00:02.172: INFO @log_profile : T read data: 1.466322 2019-10-09 11:00:02.172: INFO @log_profile : T hooks: 4.184130 2019-10-09 11:00:02.173: INFO @main_loop : Epoch 192 done 2019-10-09 11:00:02.173: INFO @main_loop : Training epoch 193 2019-10-09 11:03:13.086: INFO @log_variables: train loss mean: 0.242673 2019-10-09 11:03:13.086: INFO @log_variables: train age_loss mean: 4.068381 2019-10-09 11:03:13.086: INFO @log_variables: train gender_loss mean: 0.031952 2019-10-09 11:03:13.086: INFO @log_variables: train matching_loss nanmean: 0.313495 2019-10-09 11:03:13.086: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:03:13.086: INFO @log_variables: train age_mae mean: 4.542776 2019-10-09 11:03:13.086: INFO @log_variables: train gender_accuracy mean: 0.988993 2019-10-09 11:03:13.087: INFO @log_variables: train positive_distance nanmean: 0.740655 2019-10-09 11:03:13.087: INFO @log_variables: train negative_distance nanmean: 1.407741 2019-10-09 11:03:13.087: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:03:13.087: INFO @log_variables: valid loss mean: 0.456251 2019-10-09 11:03:13.087: INFO @log_variables: valid age_loss mean: 6.591777 2019-10-09 11:03:13.087: INFO @log_variables: valid gender_loss mean: 0.257735 2019-10-09 11:03:13.087: INFO @log_variables: valid matching_loss nanmean: 0.497466 2019-10-09 11:03:13.087: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:03:13.087: INFO @log_variables: valid age_mae mean: 7.075964 2019-10-09 11:03:13.087: INFO @log_variables: valid gender_accuracy mean: 0.931098 2019-10-09 11:03:13.087: INFO @log_variables: valid positive_distance nanmean: 0.783899 2019-10-09 11:03:13.087: INFO @log_variables: valid negative_distance nanmean: 1.375852 2019-10-09 11:03:13.087: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:03:13.087: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:03:16.222: INFO @metrics_hook: valid matching accuracy: 0.8894285542963363, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:03:16.701: INFO @decay_lr : LR updated to `0.0003800621` 2019-10-09 11:03:16.702: INFO @log_profile : T train: 180.263714 2019-10-09 11:03:16.702: INFO @log_profile : T valid: 8.508019 2019-10-09 11:03:16.702: INFO @log_profile : T read data: 1.425639 2019-10-09 11:03:16.702: INFO @log_profile : T hooks: 4.245956 2019-10-09 11:03:16.702: INFO @main_loop : Epoch 193 done 2019-10-09 11:03:16.702: INFO @main_loop : Training epoch 194 2019-10-09 11:06:26.746: INFO @log_variables: train loss mean: 0.242607 2019-10-09 11:06:26.746: INFO @log_variables: train age_loss mean: 4.068233 2019-10-09 11:06:26.746: INFO @log_variables: train gender_loss mean: 0.033330 2019-10-09 11:06:26.746: INFO @log_variables: train matching_loss nanmean: 0.311927 2019-10-09 11:06:26.746: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:06:26.746: INFO @log_variables: train age_mae mean: 4.542773 2019-10-09 11:06:26.746: INFO @log_variables: train gender_accuracy mean: 0.988061 2019-10-09 11:06:26.746: INFO @log_variables: train positive_distance nanmean: 0.738287 2019-10-09 11:06:26.746: INFO @log_variables: train negative_distance nanmean: 1.407918 2019-10-09 11:06:26.746: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:06:26.746: INFO @log_variables: valid loss mean: 0.457360 2019-10-09 11:06:26.746: INFO @log_variables: valid age_loss mean: 6.497639 2019-10-09 11:06:26.746: INFO @log_variables: valid gender_loss mean: 0.264702 2019-10-09 11:06:26.746: INFO @log_variables: valid matching_loss nanmean: 0.503350 2019-10-09 11:06:26.746: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:06:26.747: INFO @log_variables: valid age_mae mean: 6.980016 2019-10-09 11:06:26.747: INFO @log_variables: valid gender_accuracy mean: 0.927490 2019-10-09 11:06:26.747: INFO @log_variables: valid positive_distance nanmean: 0.779052 2019-10-09 11:06:26.747: INFO @log_variables: valid negative_distance nanmean: 1.372759 2019-10-09 11:06:26.747: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:06:26.747: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:06:29.145: INFO @metrics_hook: valid matching accuracy: 0.8877495952509444, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:06:29.607: INFO @decay_lr : LR updated to `0.00037816179` 2019-10-09 11:06:29.608: INFO @log_profile : T train: 179.844917 2019-10-09 11:06:29.608: INFO @log_profile : T valid: 8.505173 2019-10-09 11:06:29.608: INFO @log_profile : T read data: 1.032059 2019-10-09 11:06:29.608: INFO @log_profile : T hooks: 3.438411 2019-10-09 11:06:29.608: INFO @main_loop : Epoch 194 done 2019-10-09 11:06:29.608: INFO @main_loop : Training epoch 195 2019-10-09 11:09:40.406: INFO @log_variables: train loss mean: 0.243130 2019-10-09 11:09:40.407: INFO @log_variables: train age_loss mean: 4.070404 2019-10-09 11:09:40.407: INFO @log_variables: train gender_loss mean: 0.034985 2019-10-09 11:09:40.407: INFO @log_variables: train matching_loss nanmean: 0.311677 2019-10-09 11:09:40.407: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:09:40.407: INFO @log_variables: train age_mae mean: 4.544569 2019-10-09 11:09:40.407: INFO @log_variables: train gender_accuracy mean: 0.987962 2019-10-09 11:09:40.407: INFO @log_variables: train positive_distance nanmean: 0.740101 2019-10-09 11:09:40.407: INFO @log_variables: train negative_distance nanmean: 1.407722 2019-10-09 11:09:40.407: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:09:40.407: INFO @log_variables: valid loss mean: 0.458157 2019-10-09 11:09:40.407: INFO @log_variables: valid age_loss mean: 6.600130 2019-10-09 11:09:40.407: INFO @log_variables: valid gender_loss mean: 0.261348 2019-10-09 11:09:40.407: INFO @log_variables: valid matching_loss nanmean: 0.498926 2019-10-09 11:09:40.407: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:09:40.407: INFO @log_variables: valid age_mae mean: 7.083136 2019-10-09 11:09:40.407: INFO @log_variables: valid gender_accuracy mean: 0.924355 2019-10-09 11:09:40.408: INFO @log_variables: valid positive_distance nanmean: 0.789574 2019-10-09 11:09:40.408: INFO @log_variables: valid negative_distance nanmean: 1.376523 2019-10-09 11:09:40.408: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:09:40.408: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:09:42.489: INFO @metrics_hook: valid matching accuracy: 0.889068777358038, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:09:42.950: INFO @decay_lr : LR updated to `0.00037627097` 2019-10-09 11:09:42.951: INFO @log_profile : T train: 180.143457 2019-10-09 11:09:42.951: INFO @log_profile : T valid: 8.529136 2019-10-09 11:09:42.951: INFO @log_profile : T read data: 1.431103 2019-10-09 11:09:42.951: INFO @log_profile : T hooks: 3.152286 2019-10-09 11:09:42.951: INFO @main_loop : Epoch 195 done 2019-10-09 11:09:42.951: INFO @main_loop : Training epoch 196 2019-10-09 11:12:53.440: INFO @log_variables: train loss mean: 0.241145 2019-10-09 11:12:53.440: INFO @log_variables: train age_loss mean: 4.035896 2019-10-09 11:12:53.440: INFO @log_variables: train gender_loss mean: 0.033010 2019-10-09 11:12:53.440: INFO @log_variables: train matching_loss nanmean: 0.310950 2019-10-09 11:12:53.440: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:12:53.440: INFO @log_variables: train age_mae mean: 4.509765 2019-10-09 11:12:53.440: INFO @log_variables: train gender_accuracy mean: 0.988442 2019-10-09 11:12:53.440: INFO @log_variables: train positive_distance nanmean: 0.738793 2019-10-09 11:12:53.440: INFO @log_variables: train negative_distance nanmean: 1.407718 2019-10-09 11:12:53.440: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:12:53.440: INFO @log_variables: valid loss mean: 0.453808 2019-10-09 11:12:53.440: INFO @log_variables: valid age_loss mean: 6.442217 2019-10-09 11:12:53.440: INFO @log_variables: valid gender_loss mean: 0.266426 2019-10-09 11:12:53.441: INFO @log_variables: valid matching_loss nanmean: 0.496157 2019-10-09 11:12:53.441: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:12:53.441: INFO @log_variables: valid age_mae mean: 6.925429 2019-10-09 11:12:53.441: INFO @log_variables: valid gender_accuracy mean: 0.928732 2019-10-09 11:12:53.441: INFO @log_variables: valid positive_distance nanmean: 0.784310 2019-10-09 11:12:53.441: INFO @log_variables: valid negative_distance nanmean: 1.377015 2019-10-09 11:12:53.441: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:12:53.441: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:12:56.293: INFO @metrics_hook: valid matching accuracy: 0.8894285542963363, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:12:56.772: INFO @decay_lr : LR updated to `0.0003743896` 2019-10-09 11:12:56.773: INFO @log_profile : T train: 179.862618 2019-10-09 11:12:56.773: INFO @log_profile : T valid: 8.528668 2019-10-09 11:12:56.773: INFO @log_profile : T read data: 1.434593 2019-10-09 11:12:56.773: INFO @log_profile : T hooks: 3.910155 2019-10-09 11:12:56.773: INFO @main_loop : Epoch 196 done 2019-10-09 11:12:56.774: INFO @main_loop : Training epoch 197 2019-10-09 11:16:06.931: INFO @log_variables: train loss mean: 0.241133 2019-10-09 11:16:06.932: INFO @log_variables: train age_loss mean: 4.045889 2019-10-09 11:16:06.932: INFO @log_variables: train gender_loss mean: 0.031853 2019-10-09 11:16:06.932: INFO @log_variables: train matching_loss nanmean: 0.311070 2019-10-09 11:16:06.932: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:16:06.932: INFO @log_variables: train age_mae mean: 4.519770 2019-10-09 11:16:06.932: INFO @log_variables: train gender_accuracy mean: 0.989029 2019-10-09 11:16:06.932: INFO @log_variables: train positive_distance nanmean: 0.740004 2019-10-09 11:16:06.932: INFO @log_variables: train negative_distance nanmean: 1.408346 2019-10-09 11:16:06.932: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:16:06.932: INFO @log_variables: valid loss mean: 0.449199 2019-10-09 11:16:06.932: INFO @log_variables: valid age_loss mean: 6.297454 2019-10-09 11:16:06.932: INFO @log_variables: valid gender_loss mean: 0.263090 2019-10-09 11:16:06.932: INFO @log_variables: valid matching_loss nanmean: 0.499680 2019-10-09 11:16:06.932: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:16:06.932: INFO @log_variables: valid age_mae mean: 6.779274 2019-10-09 11:16:06.933: INFO @log_variables: valid gender_accuracy mean: 0.929915 2019-10-09 11:16:06.933: INFO @log_variables: valid positive_distance nanmean: 0.780746 2019-10-09 11:16:06.933: INFO @log_variables: valid negative_distance nanmean: 1.375433 2019-10-09 11:16:06.933: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:16:06.933: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:16:09.224: INFO @metrics_hook: valid matching accuracy: 0.8911075133417281, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:16:09.718: INFO @decay_lr : LR updated to `0.00037251765` 2019-10-09 11:16:09.719: INFO @log_profile : T train: 180.021846 2019-10-09 11:16:09.720: INFO @log_profile : T valid: 8.450484 2019-10-09 11:16:09.720: INFO @log_profile : T read data: 1.005002 2019-10-09 11:16:09.720: INFO @log_profile : T hooks: 3.383013 2019-10-09 11:16:09.720: INFO @main_loop : Epoch 197 done 2019-10-09 11:16:09.720: INFO @main_loop : Training epoch 198 2019-10-09 11:19:20.103: INFO @log_variables: train loss mean: 0.243794 2019-10-09 11:19:20.104: INFO @log_variables: train age_loss mean: 4.051397 2019-10-09 11:19:20.104: INFO @log_variables: train gender_loss mean: 0.034758 2019-10-09 11:19:20.104: INFO @log_variables: train matching_loss nanmean: 0.315863 2019-10-09 11:19:20.104: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:19:20.104: INFO @log_variables: train age_mae mean: 4.525530 2019-10-09 11:19:20.104: INFO @log_variables: train gender_accuracy mean: 0.987880 2019-10-09 11:19:20.104: INFO @log_variables: train positive_distance nanmean: 0.741479 2019-10-09 11:19:20.104: INFO @log_variables: train negative_distance nanmean: 1.407907 2019-10-09 11:19:20.104: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:19:20.104: INFO @log_variables: valid loss mean: 0.457673 2019-10-09 11:19:20.104: INFO @log_variables: valid age_loss mean: 6.546120 2019-10-09 11:19:20.104: INFO @log_variables: valid gender_loss mean: 0.260823 2019-10-09 11:19:20.104: INFO @log_variables: valid matching_loss nanmean: 0.503353 2019-10-09 11:19:20.104: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:19:20.104: INFO @log_variables: valid age_mae mean: 7.030158 2019-10-09 11:19:20.104: INFO @log_variables: valid gender_accuracy mean: 0.928614 2019-10-09 11:19:20.105: INFO @log_variables: valid positive_distance nanmean: 0.779473 2019-10-09 11:19:20.105: INFO @log_variables: valid negative_distance nanmean: 1.372670 2019-10-09 11:19:20.105: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:19:20.105: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:19:22.502: INFO @metrics_hook: valid matching accuracy: 0.8890088145349883, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:19:22.980: INFO @decay_lr : LR updated to `0.00037065506` 2019-10-09 11:19:22.981: INFO @log_profile : T train: 179.894084 2019-10-09 11:19:22.981: INFO @log_profile : T valid: 8.460365 2019-10-09 11:19:22.981: INFO @log_profile : T read data: 1.378481 2019-10-09 11:19:22.981: INFO @log_profile : T hooks: 3.443194 2019-10-09 11:19:22.982: INFO @main_loop : Epoch 198 done 2019-10-09 11:19:22.982: INFO @main_loop : Training epoch 199 2019-10-09 11:22:32.995: INFO @log_variables: train loss mean: 0.241788 2019-10-09 11:22:32.996: INFO @log_variables: train age_loss mean: 4.041472 2019-10-09 11:22:32.996: INFO @log_variables: train gender_loss mean: 0.034805 2019-10-09 11:22:32.996: INFO @log_variables: train matching_loss nanmean: 0.310590 2019-10-09 11:22:32.996: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:22:32.996: INFO @log_variables: train age_mae mean: 4.515578 2019-10-09 11:22:32.996: INFO @log_variables: train gender_accuracy mean: 0.987691 2019-10-09 11:22:32.996: INFO @log_variables: train positive_distance nanmean: 0.740106 2019-10-09 11:22:32.996: INFO @log_variables: train negative_distance nanmean: 1.407924 2019-10-09 11:22:32.996: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:22:32.996: INFO @log_variables: valid loss mean: 0.450660 2019-10-09 11:22:32.996: INFO @log_variables: valid age_loss mean: 6.352167 2019-10-09 11:22:32.996: INFO @log_variables: valid gender_loss mean: 0.262589 2019-10-09 11:22:32.996: INFO @log_variables: valid matching_loss nanmean: 0.499241 2019-10-09 11:22:32.996: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:22:32.996: INFO @log_variables: valid age_mae mean: 6.834194 2019-10-09 11:22:32.996: INFO @log_variables: valid gender_accuracy mean: 0.928081 2019-10-09 11:22:32.997: INFO @log_variables: valid positive_distance nanmean: 0.789877 2019-10-09 11:22:32.997: INFO @log_variables: valid negative_distance nanmean: 1.379465 2019-10-09 11:22:32.997: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:22:32.997: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:22:35.458: INFO @metrics_hook: valid matching accuracy: 0.8891887030041374, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:22:35.921: INFO @decay_lr : LR updated to `0.0003688018` 2019-10-09 11:22:35.922: INFO @log_profile : T train: 179.788937 2019-10-09 11:22:35.923: INFO @log_profile : T valid: 8.531985 2019-10-09 11:22:35.923: INFO @log_profile : T read data: 1.011638 2019-10-09 11:22:35.923: INFO @log_profile : T hooks: 3.522604 2019-10-09 11:22:35.923: INFO @main_loop : Epoch 199 done 2019-10-09 11:22:35.923: INFO @main_loop : Training epoch 200 2019-10-09 11:25:46.412: INFO @log_variables: train loss mean: 0.241460 2019-10-09 11:25:46.412: INFO @log_variables: train age_loss mean: 4.040783 2019-10-09 11:25:46.412: INFO @log_variables: train gender_loss mean: 0.031192 2019-10-09 11:25:46.412: INFO @log_variables: train matching_loss nanmean: 0.313257 2019-10-09 11:25:46.412: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:25:46.413: INFO @log_variables: train age_mae mean: 4.515110 2019-10-09 11:25:46.413: INFO @log_variables: train gender_accuracy mean: 0.989136 2019-10-09 11:25:46.413: INFO @log_variables: train positive_distance nanmean: 0.741056 2019-10-09 11:25:46.413: INFO @log_variables: train negative_distance nanmean: 1.407540 2019-10-09 11:25:46.413: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:25:46.413: INFO @log_variables: valid loss mean: 0.451153 2019-10-09 11:25:46.413: INFO @log_variables: valid age_loss mean: 6.333240 2019-10-09 11:25:46.413: INFO @log_variables: valid gender_loss mean: 0.267378 2019-10-09 11:25:46.413: INFO @log_variables: valid matching_loss nanmean: 0.497871 2019-10-09 11:25:46.413: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:25:46.413: INFO @log_variables: valid age_mae mean: 6.816206 2019-10-09 11:25:46.413: INFO @log_variables: valid gender_accuracy mean: 0.926071 2019-10-09 11:25:46.413: INFO @log_variables: valid positive_distance nanmean: 0.783146 2019-10-09 11:25:46.413: INFO @log_variables: valid negative_distance nanmean: 1.375772 2019-10-09 11:25:46.413: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:25:46.413: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:25:48.863: INFO @metrics_hook: valid matching accuracy: 0.8891887030041374, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:25:48.907: INFO @metrics_hook: valid age_mae: 6.816 +-0.093 (16908) 2019-10-09 11:25:48.910: INFO @metrics_hook: valid gender_accuracy: 0.926 +-0.004 (16908) 2019-10-09 11:25:50.543: INFO @decay_lr : LR updated to `0.00036695777` 2019-10-09 11:25:50.544: INFO @log_profile : T train: 179.911279 2019-10-09 11:25:50.544: INFO @log_profile : T valid: 8.513477 2019-10-09 11:25:50.544: INFO @log_profile : T read data: 1.401149 2019-10-09 11:25:50.544: INFO @log_profile : T hooks: 4.712028 2019-10-09 11:25:50.545: INFO @main_loop : Epoch 200 done 2019-10-09 11:25:50.545: INFO @main_loop : Training epoch 201 2019-10-09 11:28:59.313: INFO @log_variables: train loss mean: 0.243497 2019-10-09 11:28:59.313: INFO @log_variables: train age_loss mean: 4.068118 2019-10-09 11:28:59.313: INFO @log_variables: train gender_loss mean: 0.034757 2019-10-09 11:28:59.313: INFO @log_variables: train matching_loss nanmean: 0.313273 2019-10-09 11:28:59.313: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:28:59.313: INFO @log_variables: train age_mae mean: 4.542073 2019-10-09 11:28:59.313: INFO @log_variables: train gender_accuracy mean: 0.987898 2019-10-09 11:28:59.313: INFO @log_variables: train positive_distance nanmean: 0.738957 2019-10-09 11:28:59.313: INFO @log_variables: train negative_distance nanmean: 1.407652 2019-10-09 11:28:59.313: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:28:59.313: INFO @log_variables: valid loss mean: 0.456183 2019-10-09 11:28:59.314: INFO @log_variables: valid age_loss mean: 6.403921 2019-10-09 11:28:59.314: INFO @log_variables: valid gender_loss mean: 0.275682 2019-10-09 11:28:59.314: INFO @log_variables: valid matching_loss nanmean: 0.498093 2019-10-09 11:28:59.314: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:28:59.314: INFO @log_variables: valid age_mae mean: 6.885301 2019-10-09 11:28:59.314: INFO @log_variables: valid gender_accuracy mean: 0.924296 2019-10-09 11:28:59.314: INFO @log_variables: valid positive_distance nanmean: 0.784453 2019-10-09 11:28:59.314: INFO @log_variables: valid negative_distance nanmean: 1.377337 2019-10-09 11:28:59.314: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:28:59.314: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:29:01.840: INFO @metrics_hook: valid matching accuracy: 0.8888888888888888, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:29:02.295: INFO @decay_lr : LR updated to `0.00036512298` 2019-10-09 11:29:02.297: INFO @log_profile : T train: 178.296395 2019-10-09 11:29:02.297: INFO @log_profile : T valid: 8.354248 2019-10-09 11:29:02.297: INFO @log_profile : T read data: 1.448546 2019-10-09 11:29:02.297: INFO @log_profile : T hooks: 3.566545 2019-10-09 11:29:02.297: INFO @main_loop : Epoch 201 done 2019-10-09 11:29:02.297: INFO @main_loop : Training epoch 202 2019-10-09 11:32:10.594: INFO @log_variables: train loss mean: 0.240634 2019-10-09 11:32:10.595: INFO @log_variables: train age_loss mean: 4.030259 2019-10-09 11:32:10.595: INFO @log_variables: train gender_loss mean: 0.032876 2019-10-09 11:32:10.595: INFO @log_variables: train matching_loss nanmean: 0.310062 2019-10-09 11:32:10.595: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:32:10.595: INFO @log_variables: train age_mae mean: 4.504685 2019-10-09 11:32:10.595: INFO @log_variables: train gender_accuracy mean: 0.988604 2019-10-09 11:32:10.595: INFO @log_variables: train positive_distance nanmean: 0.738011 2019-10-09 11:32:10.595: INFO @log_variables: train negative_distance nanmean: 1.407604 2019-10-09 11:32:10.595: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:32:10.595: INFO @log_variables: valid loss mean: 0.458870 2019-10-09 11:32:10.595: INFO @log_variables: valid age_loss mean: 6.439832 2019-10-09 11:32:10.595: INFO @log_variables: valid gender_loss mean: 0.278136 2019-10-09 11:32:10.595: INFO @log_variables: valid matching_loss nanmean: 0.500379 2019-10-09 11:32:10.595: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:32:10.595: INFO @log_variables: valid age_mae mean: 6.922672 2019-10-09 11:32:10.595: INFO @log_variables: valid gender_accuracy mean: 0.925124 2019-10-09 11:32:10.595: INFO @log_variables: valid positive_distance nanmean: 0.781725 2019-10-09 11:32:10.595: INFO @log_variables: valid negative_distance nanmean: 1.376136 2019-10-09 11:32:10.595: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:32:10.596: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:32:12.878: INFO @metrics_hook: valid matching accuracy: 0.889488517119386, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:32:13.338: INFO @decay_lr : LR updated to `0.00036329735` 2019-10-09 11:32:13.339: INFO @log_profile : T train: 178.259744 2019-10-09 11:32:13.339: INFO @log_profile : T valid: 8.324062 2019-10-09 11:32:13.339: INFO @log_profile : T read data: 1.047600 2019-10-09 11:32:13.339: INFO @log_profile : T hooks: 3.324473 2019-10-09 11:32:13.339: INFO @main_loop : Epoch 202 done 2019-10-09 11:32:13.339: INFO @main_loop : Training epoch 203 2019-10-09 11:35:22.121: INFO @log_variables: train loss mean: 0.241512 2019-10-09 11:35:22.121: INFO @log_variables: train age_loss mean: 4.052827 2019-10-09 11:35:22.121: INFO @log_variables: train gender_loss mean: 0.033631 2019-10-09 11:35:22.121: INFO @log_variables: train matching_loss nanmean: 0.309775 2019-10-09 11:35:22.121: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:35:22.121: INFO @log_variables: train age_mae mean: 4.526797 2019-10-09 11:35:22.121: INFO @log_variables: train gender_accuracy mean: 0.988114 2019-10-09 11:35:22.121: INFO @log_variables: train positive_distance nanmean: 0.739025 2019-10-09 11:35:22.121: INFO @log_variables: train negative_distance nanmean: 1.407919 2019-10-09 11:35:22.121: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:35:22.121: INFO @log_variables: valid loss mean: 0.455551 2019-10-09 11:35:22.121: INFO @log_variables: valid age_loss mean: 6.553569 2019-10-09 11:35:22.121: INFO @log_variables: valid gender_loss mean: 0.259767 2019-10-09 11:35:22.122: INFO @log_variables: valid matching_loss nanmean: 0.497085 2019-10-09 11:35:22.122: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:35:22.122: INFO @log_variables: valid age_mae mean: 7.036961 2019-10-09 11:35:22.122: INFO @log_variables: valid gender_accuracy mean: 0.928436 2019-10-09 11:35:22.122: INFO @log_variables: valid positive_distance nanmean: 0.786305 2019-10-09 11:35:22.122: INFO @log_variables: valid negative_distance nanmean: 1.379150 2019-10-09 11:35:22.122: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:35:22.122: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:35:24.586: INFO @metrics_hook: valid matching accuracy: 0.8870900041973976, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:35:25.042: INFO @decay_lr : LR updated to `0.00036148087` 2019-10-09 11:35:25.043: INFO @log_profile : T train: 178.018275 2019-10-09 11:35:25.043: INFO @log_profile : T valid: 8.322386 2019-10-09 11:35:25.043: INFO @log_profile : T read data: 1.770723 2019-10-09 11:35:25.043: INFO @log_profile : T hooks: 3.508125 2019-10-09 11:35:25.043: INFO @main_loop : Epoch 203 done 2019-10-09 11:35:25.043: INFO @main_loop : Training epoch 204 2019-10-09 11:38:33.392: INFO @log_variables: train loss mean: 0.242102 2019-10-09 11:38:33.392: INFO @log_variables: train age_loss mean: 4.042717 2019-10-09 11:38:33.392: INFO @log_variables: train gender_loss mean: 0.034025 2019-10-09 11:38:33.392: INFO @log_variables: train matching_loss nanmean: 0.312219 2019-10-09 11:38:33.392: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:38:33.392: INFO @log_variables: train age_mae mean: 4.517153 2019-10-09 11:38:33.392: INFO @log_variables: train gender_accuracy mean: 0.987942 2019-10-09 11:38:33.392: INFO @log_variables: train positive_distance nanmean: 0.740315 2019-10-09 11:38:33.392: INFO @log_variables: train negative_distance nanmean: 1.407436 2019-10-09 11:38:33.392: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:38:33.392: INFO @log_variables: valid loss mean: 0.451243 2019-10-09 11:38:33.393: INFO @log_variables: valid age_loss mean: 6.366925 2019-10-09 11:38:33.393: INFO @log_variables: valid gender_loss mean: 0.264055 2019-10-09 11:38:33.393: INFO @log_variables: valid matching_loss nanmean: 0.498106 2019-10-09 11:38:33.393: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:38:33.393: INFO @log_variables: valid age_mae mean: 6.849402 2019-10-09 11:38:33.393: INFO @log_variables: valid gender_accuracy mean: 0.928377 2019-10-09 11:38:33.393: INFO @log_variables: valid positive_distance nanmean: 0.787024 2019-10-09 11:38:33.393: INFO @log_variables: valid negative_distance nanmean: 1.376899 2019-10-09 11:38:33.393: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:38:33.393: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:38:35.906: INFO @metrics_hook: valid matching accuracy: 0.8884691491275409, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:38:36.364: INFO @decay_lr : LR updated to `0.00035967346` 2019-10-09 11:38:36.366: INFO @log_profile : T train: 178.362091 2019-10-09 11:38:36.366: INFO @log_profile : T valid: 8.320415 2019-10-09 11:38:36.366: INFO @log_profile : T read data: 1.017739 2019-10-09 11:38:36.366: INFO @log_profile : T hooks: 3.534886 2019-10-09 11:38:36.366: INFO @main_loop : Epoch 204 done 2019-10-09 11:38:36.366: INFO @main_loop : Training epoch 205 2019-10-09 11:41:44.907: INFO @log_variables: train loss mean: 0.239928 2019-10-09 11:41:44.907: INFO @log_variables: train age_loss mean: 4.022734 2019-10-09 11:41:44.907: INFO @log_variables: train gender_loss mean: 0.031320 2019-10-09 11:41:44.907: INFO @log_variables: train matching_loss nanmean: 0.310185 2019-10-09 11:41:44.907: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:41:44.907: INFO @log_variables: train age_mae mean: 4.497047 2019-10-09 11:41:44.907: INFO @log_variables: train gender_accuracy mean: 0.989129 2019-10-09 11:41:44.907: INFO @log_variables: train positive_distance nanmean: 0.738652 2019-10-09 11:41:44.907: INFO @log_variables: train negative_distance nanmean: 1.407342 2019-10-09 11:41:44.907: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:41:44.907: INFO @log_variables: valid loss mean: 0.453227 2019-10-09 11:41:44.907: INFO @log_variables: valid age_loss mean: 6.398633 2019-10-09 11:41:44.907: INFO @log_variables: valid gender_loss mean: 0.268129 2019-10-09 11:41:44.908: INFO @log_variables: valid matching_loss nanmean: 0.497012 2019-10-09 11:41:44.908: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:41:44.908: INFO @log_variables: valid age_mae mean: 6.880923 2019-10-09 11:41:44.908: INFO @log_variables: valid gender_accuracy mean: 0.929205 2019-10-09 11:41:44.908: INFO @log_variables: valid positive_distance nanmean: 0.787306 2019-10-09 11:41:44.908: INFO @log_variables: valid negative_distance nanmean: 1.378865 2019-10-09 11:41:44.908: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:41:44.908: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:41:47.369: INFO @metrics_hook: valid matching accuracy: 0.8895484799424357, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:41:47.818: INFO @decay_lr : LR updated to `0.0003578751` 2019-10-09 11:41:47.819: INFO @log_profile : T train: 178.105998 2019-10-09 11:41:47.820: INFO @log_profile : T valid: 8.350697 2019-10-09 11:41:47.820: INFO @log_profile : T read data: 1.428691 2019-10-09 11:41:47.820: INFO @log_profile : T hooks: 3.483464 2019-10-09 11:41:47.820: INFO @main_loop : Epoch 205 done 2019-10-09 11:41:47.820: INFO @main_loop : Training epoch 206 2019-10-09 11:44:56.737: INFO @log_variables: train loss mean: 0.240957 2019-10-09 11:44:56.737: INFO @log_variables: train age_loss mean: 4.016943 2019-10-09 11:44:56.737: INFO @log_variables: train gender_loss mean: 0.034170 2019-10-09 11:44:56.737: INFO @log_variables: train matching_loss nanmean: 0.311101 2019-10-09 11:44:56.737: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:44:56.737: INFO @log_variables: train age_mae mean: 4.490783 2019-10-09 11:44:56.737: INFO @log_variables: train gender_accuracy mean: 0.987582 2019-10-09 11:44:56.737: INFO @log_variables: train positive_distance nanmean: 0.739409 2019-10-09 11:44:56.738: INFO @log_variables: train negative_distance nanmean: 1.407431 2019-10-09 11:44:56.738: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:44:56.738: INFO @log_variables: valid loss mean: 0.455950 2019-10-09 11:44:56.738: INFO @log_variables: valid age_loss mean: 6.494535 2019-10-09 11:44:56.738: INFO @log_variables: valid gender_loss mean: 0.266654 2019-10-09 11:44:56.738: INFO @log_variables: valid matching_loss nanmean: 0.497338 2019-10-09 11:44:56.738: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:44:56.738: INFO @log_variables: valid age_mae mean: 6.977983 2019-10-09 11:44:56.738: INFO @log_variables: valid gender_accuracy mean: 0.926839 2019-10-09 11:44:56.738: INFO @log_variables: valid positive_distance nanmean: 0.786604 2019-10-09 11:44:56.738: INFO @log_variables: valid negative_distance nanmean: 1.377228 2019-10-09 11:44:56.738: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:44:56.738: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:44:58.838: INFO @metrics_hook: valid matching accuracy: 0.8906877735803802, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:44:59.290: INFO @decay_lr : LR updated to `0.00035608574` 2019-10-09 11:44:59.291: INFO @log_profile : T train: 178.538015 2019-10-09 11:44:59.291: INFO @log_profile : T valid: 8.291950 2019-10-09 11:44:59.291: INFO @log_profile : T read data: 1.424680 2019-10-09 11:44:59.291: INFO @log_profile : T hooks: 3.130079 2019-10-09 11:44:59.291: INFO @main_loop : Epoch 206 done 2019-10-09 11:44:59.291: INFO @main_loop : Training epoch 207 2019-10-09 11:48:07.554: INFO @log_variables: train loss mean: 0.241813 2019-10-09 11:48:07.554: INFO @log_variables: train age_loss mean: 4.047980 2019-10-09 11:48:07.554: INFO @log_variables: train gender_loss mean: 0.033027 2019-10-09 11:48:07.554: INFO @log_variables: train matching_loss nanmean: 0.311797 2019-10-09 11:48:07.554: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:48:07.554: INFO @log_variables: train age_mae mean: 4.522006 2019-10-09 11:48:07.554: INFO @log_variables: train gender_accuracy mean: 0.988585 2019-10-09 11:48:07.554: INFO @log_variables: train positive_distance nanmean: 0.739515 2019-10-09 11:48:07.554: INFO @log_variables: train negative_distance nanmean: 1.407464 2019-10-09 11:48:07.554: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:48:07.554: INFO @log_variables: valid loss mean: 0.444062 2019-10-09 11:48:07.554: INFO @log_variables: valid age_loss mean: 6.287663 2019-10-09 11:48:07.554: INFO @log_variables: valid gender_loss mean: 0.253417 2019-10-09 11:48:07.554: INFO @log_variables: valid matching_loss nanmean: 0.494409 2019-10-09 11:48:07.555: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:48:07.555: INFO @log_variables: valid age_mae mean: 6.770096 2019-10-09 11:48:07.555: INFO @log_variables: valid gender_accuracy mean: 0.930920 2019-10-09 11:48:07.555: INFO @log_variables: valid positive_distance nanmean: 0.784521 2019-10-09 11:48:07.555: INFO @log_variables: valid negative_distance nanmean: 1.377800 2019-10-09 11:48:07.555: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:48:07.555: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:48:09.983: INFO @metrics_hook: valid matching accuracy: 0.890507885111231, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:48:10.425: INFO @decay_lr : LR updated to `0.00035430532` 2019-10-09 11:48:10.426: INFO @log_profile : T train: 178.184269 2019-10-09 11:48:10.426: INFO @log_profile : T valid: 8.366442 2019-10-09 11:48:10.426: INFO @log_profile : T read data: 1.036864 2019-10-09 11:48:10.426: INFO @log_profile : T hooks: 3.461636 2019-10-09 11:48:10.426: INFO @main_loop : Epoch 207 done 2019-10-09 11:48:10.426: INFO @main_loop : Training epoch 208 2019-10-09 11:51:19.162: INFO @log_variables: train loss mean: 0.239132 2019-10-09 11:51:19.162: INFO @log_variables: train age_loss mean: 4.001011 2019-10-09 11:51:19.163: INFO @log_variables: train gender_loss mean: 0.031991 2019-10-09 11:51:19.163: INFO @log_variables: train matching_loss nanmean: 0.309218 2019-10-09 11:51:19.163: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:51:19.163: INFO @log_variables: train age_mae mean: 4.474563 2019-10-09 11:51:19.163: INFO @log_variables: train gender_accuracy mean: 0.988740 2019-10-09 11:51:19.163: INFO @log_variables: train positive_distance nanmean: 0.738112 2019-10-09 11:51:19.163: INFO @log_variables: train negative_distance nanmean: 1.407577 2019-10-09 11:51:19.163: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:51:19.163: INFO @log_variables: valid loss mean: 0.459794 2019-10-09 11:51:19.163: INFO @log_variables: valid age_loss mean: 6.618663 2019-10-09 11:51:19.163: INFO @log_variables: valid gender_loss mean: 0.262167 2019-10-09 11:51:19.163: INFO @log_variables: valid matching_loss nanmean: 0.501327 2019-10-09 11:51:19.163: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:51:19.163: INFO @log_variables: valid age_mae mean: 7.102292 2019-10-09 11:51:19.163: INFO @log_variables: valid gender_accuracy mean: 0.929383 2019-10-09 11:51:19.163: INFO @log_variables: valid positive_distance nanmean: 0.779420 2019-10-09 11:51:19.163: INFO @log_variables: valid negative_distance nanmean: 1.373201 2019-10-09 11:51:19.163: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:51:19.164: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:51:21.328: INFO @metrics_hook: valid matching accuracy: 0.8899682197037837, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:51:21.804: INFO @decay_lr : LR updated to `0.0003525338` 2019-10-09 11:51:21.806: INFO @log_profile : T train: 178.226354 2019-10-09 11:51:21.806: INFO @log_profile : T valid: 8.366839 2019-10-09 11:51:21.806: INFO @log_profile : T read data: 1.478070 2019-10-09 11:51:21.806: INFO @log_profile : T hooks: 3.221778 2019-10-09 11:51:21.806: INFO @main_loop : Epoch 208 done 2019-10-09 11:51:21.806: INFO @main_loop : Training epoch 209 2019-10-09 11:54:30.457: INFO @log_variables: train loss mean: 0.240878 2019-10-09 11:54:30.457: INFO @log_variables: train age_loss mean: 4.016553 2019-10-09 11:54:30.457: INFO @log_variables: train gender_loss mean: 0.033468 2019-10-09 11:54:30.457: INFO @log_variables: train matching_loss nanmean: 0.311598 2019-10-09 11:54:30.458: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:54:30.458: INFO @log_variables: train age_mae mean: 4.490468 2019-10-09 11:54:30.458: INFO @log_variables: train gender_accuracy mean: 0.988441 2019-10-09 11:54:30.458: INFO @log_variables: train positive_distance nanmean: 0.738488 2019-10-09 11:54:30.458: INFO @log_variables: train negative_distance nanmean: 1.407427 2019-10-09 11:54:30.458: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:54:30.458: INFO @log_variables: valid loss mean: 0.453179 2019-10-09 11:54:30.458: INFO @log_variables: valid age_loss mean: 6.426670 2019-10-09 11:54:30.458: INFO @log_variables: valid gender_loss mean: 0.261837 2019-10-09 11:54:30.458: INFO @log_variables: valid matching_loss nanmean: 0.500352 2019-10-09 11:54:30.458: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:54:30.458: INFO @log_variables: valid age_mae mean: 6.909902 2019-10-09 11:54:30.458: INFO @log_variables: valid gender_accuracy mean: 0.928377 2019-10-09 11:54:30.458: INFO @log_variables: valid positive_distance nanmean: 0.784779 2019-10-09 11:54:30.459: INFO @log_variables: valid negative_distance nanmean: 1.376365 2019-10-09 11:54:30.459: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:54:30.459: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:54:33.009: INFO @metrics_hook: valid matching accuracy: 0.8885291119505906, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:54:33.458: INFO @decay_lr : LR updated to `0.00035077112` 2019-10-09 11:54:33.459: INFO @log_profile : T train: 178.263525 2019-10-09 11:54:33.459: INFO @log_profile : T valid: 8.351370 2019-10-09 11:54:33.459: INFO @log_profile : T read data: 1.374800 2019-10-09 11:54:33.459: INFO @log_profile : T hooks: 3.579171 2019-10-09 11:54:33.459: INFO @main_loop : Epoch 209 done 2019-10-09 11:54:33.459: INFO @main_loop : Training epoch 210 2019-10-09 11:57:41.579: INFO @log_variables: train loss mean: 0.238866 2019-10-09 11:57:41.580: INFO @log_variables: train age_loss mean: 4.000082 2019-10-09 11:57:41.580: INFO @log_variables: train gender_loss mean: 0.031048 2019-10-09 11:57:41.580: INFO @log_variables: train matching_loss nanmean: 0.309429 2019-10-09 11:57:41.580: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 11:57:41.580: INFO @log_variables: train age_mae mean: 4.474144 2019-10-09 11:57:41.580: INFO @log_variables: train gender_accuracy mean: 0.989190 2019-10-09 11:57:41.580: INFO @log_variables: train positive_distance nanmean: 0.738005 2019-10-09 11:57:41.580: INFO @log_variables: train negative_distance nanmean: 1.407459 2019-10-09 11:57:41.580: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 11:57:41.580: INFO @log_variables: valid loss mean: 0.456238 2019-10-09 11:57:41.580: INFO @log_variables: valid age_loss mean: 6.403634 2019-10-09 11:57:41.580: INFO @log_variables: valid gender_loss mean: 0.273777 2019-10-09 11:57:41.580: INFO @log_variables: valid matching_loss nanmean: 0.500197 2019-10-09 11:57:41.580: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 11:57:41.580: INFO @log_variables: valid age_mae mean: 6.886516 2019-10-09 11:57:41.580: INFO @log_variables: valid gender_accuracy mean: 0.927667 2019-10-09 11:57:41.580: INFO @log_variables: valid positive_distance nanmean: 0.783556 2019-10-09 11:57:41.581: INFO @log_variables: valid negative_distance nanmean: 1.377228 2019-10-09 11:57:41.581: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 11:57:41.581: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 11:57:43.468: INFO @metrics_hook: valid matching accuracy: 0.8903879594651316, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 11:57:43.922: INFO @decay_lr : LR updated to `0.00034901727` 2019-10-09 11:57:43.924: INFO @log_profile : T train: 178.023409 2019-10-09 11:57:43.924: INFO @log_profile : T valid: 8.399051 2019-10-09 11:57:43.924: INFO @log_profile : T read data: 1.033058 2019-10-09 11:57:43.924: INFO @log_profile : T hooks: 2.923311 2019-10-09 11:57:43.924: INFO @main_loop : Epoch 210 done 2019-10-09 11:57:43.924: INFO @main_loop : Training epoch 211 2019-10-09 12:00:52.545: INFO @log_variables: train loss mean: 0.239638 2019-10-09 12:00:52.545: INFO @log_variables: train age_loss mean: 4.025121 2019-10-09 12:00:52.545: INFO @log_variables: train gender_loss mean: 0.032405 2019-10-09 12:00:52.545: INFO @log_variables: train matching_loss nanmean: 0.307961 2019-10-09 12:00:52.545: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:00:52.545: INFO @log_variables: train age_mae mean: 4.498937 2019-10-09 12:00:52.545: INFO @log_variables: train gender_accuracy mean: 0.988928 2019-10-09 12:00:52.545: INFO @log_variables: train positive_distance nanmean: 0.737180 2019-10-09 12:00:52.545: INFO @log_variables: train negative_distance nanmean: 1.407540 2019-10-09 12:00:52.545: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:00:52.545: INFO @log_variables: valid loss mean: 0.454698 2019-10-09 12:00:52.545: INFO @log_variables: valid age_loss mean: 6.495241 2019-10-09 12:00:52.545: INFO @log_variables: valid gender_loss mean: 0.260821 2019-10-09 12:00:52.545: INFO @log_variables: valid matching_loss nanmean: 0.499221 2019-10-09 12:00:52.545: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:00:52.545: INFO @log_variables: valid age_mae mean: 6.978042 2019-10-09 12:00:52.545: INFO @log_variables: valid gender_accuracy mean: 0.932281 2019-10-09 12:00:52.546: INFO @log_variables: valid positive_distance nanmean: 0.783921 2019-10-09 12:00:52.546: INFO @log_variables: valid negative_distance nanmean: 1.376576 2019-10-09 12:00:52.546: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:00:52.546: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:00:54.841: INFO @metrics_hook: valid matching accuracy: 0.8881093721892427, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:00:55.296: INFO @decay_lr : LR updated to `0.00034727217` 2019-10-09 12:00:55.995: INFO @model : Quantizing and saving the model 2019-10-09 12:00:56.825: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/mul 2019-10-09 12:00:56.831: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_0/bn-device-gpu-0/batchnorm/add_1 2019-10-09 12:00:56.837: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/mul 2019-10-09 12:00:56.842: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_1/bn-device-gpu-0/batchnorm/add_1 2019-10-09 12:00:56.848: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/mul 2019-10-09 12:00:56.853: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_2/bn-device-gpu-0/batchnorm/add_1 2019-10-09 12:00:56.859: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/mul 2019-10-09 12:00:56.864: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_4/bn-device-gpu-0/batchnorm/add_1 2019-10-09 12:00:56.870: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/mul 2019-10-09 12:00:56.875: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_5/bn-device-gpu-0/batchnorm/add_1 2019-10-09 12:00:56.881: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/mul 2019-10-09 12:00:56.886: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_6/bn-device-gpu-0/batchnorm/add_1 2019-10-09 12:00:56.893: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/mul 2019-10-09 12:00:56.898: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_8/bn-device-gpu-0/batchnorm/add_1 2019-10-09 12:00:56.903: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/mul 2019-10-09 12:00:56.909: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_9/bn-device-gpu-0/batchnorm/add_1 2019-10-09 12:00:56.914: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/mul 2019-10-09 12:00:56.920: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_10/bn-device-gpu-0/batchnorm/add_1 2019-10-09 12:00:56.925: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/mul 2019-10-09 12:00:56.930: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_12/bn-device-gpu-0/batchnorm/add_1 2019-10-09 12:00:56.935: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/mul 2019-10-09 12:00:56.940: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_13/bn-device-gpu-0/batchnorm/add_1 2019-10-09 12:00:56.945: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/mul 2019-10-09 12:00:56.950: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_14/bn-device-gpu-0/batchnorm/add_1 2019-10-09 12:00:56.955: INFO @quantize : Inserting fake quant op activation_Mul_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/mul 2019-10-09 12:00:56.960: INFO @quantize : Inserting fake quant op activation_Add_quant after main_branch/ConvBlock_16/bn-device-gpu-0/batchnorm/add_1 2019-10-09 12:00:56.965: INFO @quantize : Inserting fake quant op activation_Mul_quant after matching_branch/l2_normalize 2019-10-09 12:00:56.973: INFO @model : Copying variables to the eval graph, this may take a while 2019-10-09 12:01:09.764: INFO @model : Converting model with inputs [] and outputs f[, , ] 2019-10-09 12:01:10.051: INFO @graph_util_impl: Froze 195 variables. 2019-10-09 12:01:10.070: INFO @graph_util_impl: Converted 195 variables to const ops. 2019-10-09 12:01:12.088: INFO @model : Saving the quantized model to `./log/RecognitionQuantized_suspicious-leavitt/quantized.tflite` 2019-10-09 12:01:12.133: INFO @save : Model saved to: ./log/RecognitionQuantized_suspicious-leavitt/model_best_gender.ckpt 2019-10-09 12:01:12.138: INFO @log_profile : T train: 178.239579 2019-10-09 12:01:12.138: INFO @log_profile : T valid: 8.314805 2019-10-09 12:01:12.138: INFO @log_profile : T read data: 1.398563 2019-10-09 12:01:12.138: INFO @log_profile : T hooks: 20.174032 2019-10-09 12:01:12.138: INFO @main_loop : Epoch 211 done 2019-10-09 12:01:12.138: INFO @main_loop : Training epoch 212 2019-10-09 12:04:20.350: INFO @log_variables: train loss mean: 0.237544 2019-10-09 12:04:20.350: INFO @log_variables: train age_loss mean: 3.973140 2019-10-09 12:04:20.350: INFO @log_variables: train gender_loss mean: 0.030941 2019-10-09 12:04:20.350: INFO @log_variables: train matching_loss nanmean: 0.308131 2019-10-09 12:04:20.350: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:04:20.350: INFO @log_variables: train age_mae mean: 4.446742 2019-10-09 12:04:20.350: INFO @log_variables: train gender_accuracy mean: 0.989210 2019-10-09 12:04:20.350: INFO @log_variables: train positive_distance nanmean: 0.736349 2019-10-09 12:04:20.350: INFO @log_variables: train negative_distance nanmean: 1.407602 2019-10-09 12:04:20.350: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:04:20.350: INFO @log_variables: valid loss mean: 0.454568 2019-10-09 12:04:20.350: INFO @log_variables: valid age_loss mean: 6.510332 2019-10-09 12:04:20.350: INFO @log_variables: valid gender_loss mean: 0.258084 2019-10-09 12:04:20.350: INFO @log_variables: valid matching_loss nanmean: 0.500043 2019-10-09 12:04:20.350: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:04:20.351: INFO @log_variables: valid age_mae mean: 6.993555 2019-10-09 12:04:20.351: INFO @log_variables: valid gender_accuracy mean: 0.930447 2019-10-09 12:04:20.351: INFO @log_variables: valid positive_distance nanmean: 0.782692 2019-10-09 12:04:20.351: INFO @log_variables: valid negative_distance nanmean: 1.375996 2019-10-09 12:04:20.351: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:04:20.351: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:04:22.956: INFO @metrics_hook: valid matching accuracy: 0.889068777358038, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:04:23.406: INFO @decay_lr : LR updated to `0.0003455358` 2019-10-09 12:04:23.407: INFO @log_profile : T train: 178.224538 2019-10-09 12:04:23.407: INFO @log_profile : T valid: 8.296215 2019-10-09 12:04:23.407: INFO @log_profile : T read data: 1.042738 2019-10-09 12:04:23.407: INFO @log_profile : T hooks: 3.619829 2019-10-09 12:04:23.407: INFO @main_loop : Epoch 212 done 2019-10-09 12:04:23.407: INFO @main_loop : Training epoch 213 2019-10-09 12:07:31.919: INFO @log_variables: train loss mean: 0.238677 2019-10-09 12:07:31.919: INFO @log_variables: train age_loss mean: 3.983241 2019-10-09 12:07:31.919: INFO @log_variables: train gender_loss mean: 0.031127 2019-10-09 12:07:31.919: INFO @log_variables: train matching_loss nanmean: 0.310447 2019-10-09 12:07:31.919: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:07:31.919: INFO @log_variables: train age_mae mean: 4.456916 2019-10-09 12:07:31.919: INFO @log_variables: train gender_accuracy mean: 0.989561 2019-10-09 12:07:31.919: INFO @log_variables: train positive_distance nanmean: 0.738983 2019-10-09 12:07:31.919: INFO @log_variables: train negative_distance nanmean: 1.407687 2019-10-09 12:07:31.919: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:07:31.919: INFO @log_variables: valid loss mean: 0.451484 2019-10-09 12:07:31.919: INFO @log_variables: valid age_loss mean: 6.356058 2019-10-09 12:07:31.919: INFO @log_variables: valid gender_loss mean: 0.263607 2019-10-09 12:07:31.919: INFO @log_variables: valid matching_loss nanmean: 0.500386 2019-10-09 12:07:31.919: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:07:31.919: INFO @log_variables: valid age_mae mean: 6.839657 2019-10-09 12:07:31.920: INFO @log_variables: valid gender_accuracy mean: 0.931748 2019-10-09 12:07:31.920: INFO @log_variables: valid positive_distance nanmean: 0.785961 2019-10-09 12:07:31.920: INFO @log_variables: valid negative_distance nanmean: 1.375985 2019-10-09 12:07:31.920: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:07:31.920: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:07:34.239: INFO @metrics_hook: valid matching accuracy: 0.8874497811356958, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:07:34.657: INFO @decay_lr : LR updated to `0.00034380812` 2019-10-09 12:07:34.658: INFO @log_profile : T train: 178.097241 2019-10-09 12:07:34.658: INFO @log_profile : T valid: 8.327255 2019-10-09 12:07:34.658: INFO @log_profile : T read data: 1.412118 2019-10-09 12:07:34.658: INFO @log_profile : T hooks: 3.326994 2019-10-09 12:07:34.658: INFO @main_loop : Epoch 213 done 2019-10-09 12:07:34.658: INFO @main_loop : Training epoch 214 2019-10-09 12:10:43.162: INFO @log_variables: train loss mean: 0.240111 2019-10-09 12:10:43.162: INFO @log_variables: train age_loss mean: 4.008774 2019-10-09 12:10:43.162: INFO @log_variables: train gender_loss mean: 0.031614 2019-10-09 12:10:43.162: INFO @log_variables: train matching_loss nanmean: 0.311853 2019-10-09 12:10:43.163: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:10:43.163: INFO @log_variables: train age_mae mean: 4.482939 2019-10-09 12:10:43.163: INFO @log_variables: train gender_accuracy mean: 0.988755 2019-10-09 12:10:43.163: INFO @log_variables: train positive_distance nanmean: 0.739421 2019-10-09 12:10:43.163: INFO @log_variables: train negative_distance nanmean: 1.407564 2019-10-09 12:10:43.163: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:10:43.163: INFO @log_variables: valid loss mean: 0.459749 2019-10-09 12:10:43.163: INFO @log_variables: valid age_loss mean: 6.396048 2019-10-09 12:10:43.163: INFO @log_variables: valid gender_loss mean: 0.288980 2019-10-09 12:10:43.163: INFO @log_variables: valid matching_loss nanmean: 0.496636 2019-10-09 12:10:43.163: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:10:43.163: INFO @log_variables: valid age_mae mean: 6.878746 2019-10-09 12:10:43.163: INFO @log_variables: valid gender_accuracy mean: 0.927194 2019-10-09 12:10:43.163: INFO @log_variables: valid positive_distance nanmean: 0.780702 2019-10-09 12:10:43.163: INFO @log_variables: valid negative_distance nanmean: 1.375964 2019-10-09 12:10:43.163: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:10:43.163: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:10:45.820: INFO @metrics_hook: valid matching accuracy: 0.8899682197037837, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:10:46.270: INFO @decay_lr : LR updated to `0.00034208907` 2019-10-09 12:10:46.272: INFO @log_profile : T train: 178.086782 2019-10-09 12:10:46.272: INFO @log_profile : T valid: 8.350720 2019-10-09 12:10:46.272: INFO @log_profile : T read data: 1.418299 2019-10-09 12:10:46.272: INFO @log_profile : T hooks: 3.670304 2019-10-09 12:10:46.272: INFO @main_loop : Epoch 214 done 2019-10-09 12:10:46.272: INFO @main_loop : Training epoch 215 2019-10-09 12:13:54.325: INFO @log_variables: train loss mean: 0.238669 2019-10-09 12:13:54.325: INFO @log_variables: train age_loss mean: 4.004776 2019-10-09 12:13:54.325: INFO @log_variables: train gender_loss mean: 0.029895 2019-10-09 12:13:54.325: INFO @log_variables: train matching_loss nanmean: 0.309503 2019-10-09 12:13:54.325: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:13:54.325: INFO @log_variables: train age_mae mean: 4.478749 2019-10-09 12:13:54.325: INFO @log_variables: train gender_accuracy mean: 0.989245 2019-10-09 12:13:54.325: INFO @log_variables: train positive_distance nanmean: 0.738097 2019-10-09 12:13:54.325: INFO @log_variables: train negative_distance nanmean: 1.407480 2019-10-09 12:13:54.325: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:13:54.325: INFO @log_variables: valid loss mean: 0.450770 2019-10-09 12:13:54.326: INFO @log_variables: valid age_loss mean: 6.309358 2019-10-09 12:13:54.326: INFO @log_variables: valid gender_loss mean: 0.267621 2019-10-09 12:13:54.326: INFO @log_variables: valid matching_loss nanmean: 0.498829 2019-10-09 12:13:54.326: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:13:54.326: INFO @log_variables: valid age_mae mean: 6.791772 2019-10-09 12:13:54.326: INFO @log_variables: valid gender_accuracy mean: 0.928791 2019-10-09 12:13:54.326: INFO @log_variables: valid positive_distance nanmean: 0.782908 2019-10-09 12:13:54.326: INFO @log_variables: valid negative_distance nanmean: 1.375955 2019-10-09 12:13:54.326: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:13:54.326: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:13:56.855: INFO @metrics_hook: valid matching accuracy: 0.8887689632427894, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:13:57.298: INFO @decay_lr : LR updated to `0.00034037864` 2019-10-09 12:13:57.300: INFO @log_profile : T train: 178.056476 2019-10-09 12:13:57.300: INFO @log_profile : T valid: 8.374187 2019-10-09 12:13:57.300: INFO @log_profile : T read data: 0.976024 2019-10-09 12:13:57.300: INFO @log_profile : T hooks: 3.523476 2019-10-09 12:13:57.300: INFO @main_loop : Epoch 215 done 2019-10-09 12:13:57.300: INFO @main_loop : Training epoch 216 2019-10-09 12:17:05.990: INFO @log_variables: train loss mean: 0.239538 2019-10-09 12:17:05.991: INFO @log_variables: train age_loss mean: 4.024914 2019-10-09 12:17:05.991: INFO @log_variables: train gender_loss mean: 0.031111 2019-10-09 12:17:05.991: INFO @log_variables: train matching_loss nanmean: 0.308964 2019-10-09 12:17:05.991: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:17:05.991: INFO @log_variables: train age_mae mean: 4.499318 2019-10-09 12:17:05.991: INFO @log_variables: train gender_accuracy mean: 0.989173 2019-10-09 12:17:05.991: INFO @log_variables: train positive_distance nanmean: 0.738676 2019-10-09 12:17:05.991: INFO @log_variables: train negative_distance nanmean: 1.407704 2019-10-09 12:17:05.991: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:17:05.991: INFO @log_variables: valid loss mean: 0.450354 2019-10-09 12:17:05.991: INFO @log_variables: valid age_loss mean: 6.349629 2019-10-09 12:17:05.991: INFO @log_variables: valid gender_loss mean: 0.260386 2019-10-09 12:17:05.991: INFO @log_variables: valid matching_loss nanmean: 0.500748 2019-10-09 12:17:05.991: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:17:05.991: INFO @log_variables: valid age_mae mean: 6.832341 2019-10-09 12:17:05.991: INFO @log_variables: valid gender_accuracy mean: 0.927253 2019-10-09 12:17:05.991: INFO @log_variables: valid positive_distance nanmean: 0.784912 2019-10-09 12:17:05.991: INFO @log_variables: valid negative_distance nanmean: 1.378410 2019-10-09 12:17:05.992: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:17:05.992: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:17:08.544: INFO @metrics_hook: valid matching accuracy: 0.8887689632427894, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:17:08.999: INFO @decay_lr : LR updated to `0.00033867673` 2019-10-09 12:17:09.001: INFO @log_profile : T train: 178.067362 2019-10-09 12:17:09.001: INFO @log_profile : T valid: 8.579619 2019-10-09 12:17:09.001: INFO @log_profile : T read data: 1.410413 2019-10-09 12:17:09.001: INFO @log_profile : T hooks: 3.559456 2019-10-09 12:17:09.001: INFO @main_loop : Epoch 216 done 2019-10-09 12:17:09.001: INFO @main_loop : Training epoch 217 2019-10-09 12:20:17.368: INFO @log_variables: train loss mean: 0.239298 2019-10-09 12:20:17.368: INFO @log_variables: train age_loss mean: 3.999155 2019-10-09 12:20:17.368: INFO @log_variables: train gender_loss mean: 0.031553 2019-10-09 12:20:17.368: INFO @log_variables: train matching_loss nanmean: 0.310354 2019-10-09 12:20:17.368: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:20:17.369: INFO @log_variables: train age_mae mean: 4.473091 2019-10-09 12:20:17.369: INFO @log_variables: train gender_accuracy mean: 0.989346 2019-10-09 12:20:17.369: INFO @log_variables: train positive_distance nanmean: 0.738469 2019-10-09 12:20:17.369: INFO @log_variables: train negative_distance nanmean: 1.407422 2019-10-09 12:20:17.369: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:20:17.369: INFO @log_variables: valid loss mean: 0.450675 2019-10-09 12:20:17.369: INFO @log_variables: valid age_loss mean: 6.325702 2019-10-09 12:20:17.369: INFO @log_variables: valid gender_loss mean: 0.261129 2019-10-09 12:20:17.369: INFO @log_variables: valid matching_loss nanmean: 0.503392 2019-10-09 12:20:17.369: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:20:17.369: INFO @log_variables: valid age_mae mean: 6.809226 2019-10-09 12:20:17.369: INFO @log_variables: valid gender_accuracy mean: 0.928259 2019-10-09 12:20:17.369: INFO @log_variables: valid positive_distance nanmean: 0.784429 2019-10-09 12:20:17.369: INFO @log_variables: valid negative_distance nanmean: 1.375410 2019-10-09 12:20:17.369: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:20:17.370: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:20:20.431: INFO @metrics_hook: valid matching accuracy: 0.8883492234814415, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:20:20.905: INFO @decay_lr : LR updated to `0.00033698336` 2019-10-09 12:20:20.907: INFO @log_profile : T train: 178.047450 2019-10-09 12:20:20.907: INFO @log_profile : T valid: 8.289105 2019-10-09 12:20:20.907: INFO @log_profile : T read data: 1.377042 2019-10-09 12:20:20.907: INFO @log_profile : T hooks: 4.107423 2019-10-09 12:20:20.907: INFO @main_loop : Epoch 217 done 2019-10-09 12:20:20.907: INFO @main_loop : Training epoch 218 2019-10-09 12:23:29.230: INFO @log_variables: train loss mean: 0.238353 2019-10-09 12:23:29.230: INFO @log_variables: train age_loss mean: 3.992194 2019-10-09 12:23:29.230: INFO @log_variables: train gender_loss mean: 0.030524 2019-10-09 12:23:29.230: INFO @log_variables: train matching_loss nanmean: 0.309151 2019-10-09 12:23:29.230: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:23:29.230: INFO @log_variables: train age_mae mean: 4.466329 2019-10-09 12:23:29.230: INFO @log_variables: train gender_accuracy mean: 0.989427 2019-10-09 12:23:29.230: INFO @log_variables: train positive_distance nanmean: 0.738441 2019-10-09 12:23:29.230: INFO @log_variables: train negative_distance nanmean: 1.407173 2019-10-09 12:23:29.230: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:23:29.230: INFO @log_variables: valid loss mean: 0.453197 2019-10-09 12:23:29.231: INFO @log_variables: valid age_loss mean: 6.512348 2019-10-09 12:23:29.231: INFO @log_variables: valid gender_loss mean: 0.256856 2019-10-09 12:23:29.231: INFO @log_variables: valid matching_loss nanmean: 0.496821 2019-10-09 12:23:29.231: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:23:29.231: INFO @log_variables: valid age_mae mean: 6.995118 2019-10-09 12:23:29.231: INFO @log_variables: valid gender_accuracy mean: 0.930743 2019-10-09 12:23:29.231: INFO @log_variables: valid positive_distance nanmean: 0.782604 2019-10-09 12:23:29.231: INFO @log_variables: valid negative_distance nanmean: 1.377228 2019-10-09 12:23:29.231: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:23:29.231: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:23:32.000: INFO @metrics_hook: valid matching accuracy: 0.8897283684115849, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:23:32.473: INFO @decay_lr : LR updated to `0.00033529845` 2019-10-09 12:23:32.475: INFO @log_profile : T train: 178.239270 2019-10-09 12:23:32.475: INFO @log_profile : T valid: 8.389813 2019-10-09 12:23:32.475: INFO @log_profile : T read data: 1.024415 2019-10-09 12:23:32.475: INFO @log_profile : T hooks: 3.826586 2019-10-09 12:23:32.475: INFO @main_loop : Epoch 218 done 2019-10-09 12:23:32.475: INFO @main_loop : Training epoch 219 2019-10-09 12:26:41.052: INFO @log_variables: train loss mean: 0.237729 2019-10-09 12:26:41.052: INFO @log_variables: train age_loss mean: 3.962620 2019-10-09 12:26:41.052: INFO @log_variables: train gender_loss mean: 0.030715 2019-10-09 12:26:41.052: INFO @log_variables: train matching_loss nanmean: 0.309983 2019-10-09 12:26:41.052: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:26:41.052: INFO @log_variables: train age_mae mean: 4.436465 2019-10-09 12:26:41.052: INFO @log_variables: train gender_accuracy mean: 0.989346 2019-10-09 12:26:41.053: INFO @log_variables: train positive_distance nanmean: 0.737925 2019-10-09 12:26:41.053: INFO @log_variables: train negative_distance nanmean: 1.407647 2019-10-09 12:26:41.053: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:26:41.053: INFO @log_variables: valid loss mean: 0.453029 2019-10-09 12:26:41.053: INFO @log_variables: valid age_loss mean: 6.341538 2019-10-09 12:26:41.053: INFO @log_variables: valid gender_loss mean: 0.270852 2019-10-09 12:26:41.053: INFO @log_variables: valid matching_loss nanmean: 0.499383 2019-10-09 12:26:41.053: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:26:41.053: INFO @log_variables: valid age_mae mean: 6.823911 2019-10-09 12:26:41.053: INFO @log_variables: valid gender_accuracy mean: 0.927608 2019-10-09 12:26:41.053: INFO @log_variables: valid positive_distance nanmean: 0.783185 2019-10-09 12:26:41.053: INFO @log_variables: valid negative_distance nanmean: 1.375322 2019-10-09 12:26:41.053: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:26:41.053: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:26:43.293: INFO @metrics_hook: valid matching accuracy: 0.8885291119505906, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:26:43.757: INFO @decay_lr : LR updated to `0.00033362195` 2019-10-09 12:26:43.758: INFO @log_profile : T train: 178.132050 2019-10-09 12:26:43.758: INFO @log_profile : T valid: 8.358093 2019-10-09 12:26:43.758: INFO @log_profile : T read data: 1.435345 2019-10-09 12:26:43.758: INFO @log_profile : T hooks: 3.272347 2019-10-09 12:26:43.759: INFO @main_loop : Epoch 219 done 2019-10-09 12:26:43.759: INFO @main_loop : Training epoch 220 2019-10-09 12:29:52.074: INFO @log_variables: train loss mean: 0.238630 2019-10-09 12:29:52.074: INFO @log_variables: train age_loss mean: 3.986464 2019-10-09 12:29:52.074: INFO @log_variables: train gender_loss mean: 0.032208 2019-10-09 12:29:52.075: INFO @log_variables: train matching_loss nanmean: 0.308897 2019-10-09 12:29:52.075: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:29:52.075: INFO @log_variables: train age_mae mean: 4.459764 2019-10-09 12:29:52.075: INFO @log_variables: train gender_accuracy mean: 0.989018 2019-10-09 12:29:52.075: INFO @log_variables: train positive_distance nanmean: 0.739655 2019-10-09 12:29:52.075: INFO @log_variables: train negative_distance nanmean: 1.407585 2019-10-09 12:29:52.075: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:29:52.075: INFO @log_variables: valid loss mean: 0.453542 2019-10-09 12:29:52.075: INFO @log_variables: valid age_loss mean: 6.282324 2019-10-09 12:29:52.075: INFO @log_variables: valid gender_loss mean: 0.280265 2019-10-09 12:29:52.075: INFO @log_variables: valid matching_loss nanmean: 0.497484 2019-10-09 12:29:52.075: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:29:52.075: INFO @log_variables: valid age_mae mean: 6.765293 2019-10-09 12:29:52.075: INFO @log_variables: valid gender_accuracy mean: 0.928555 2019-10-09 12:29:52.075: INFO @log_variables: valid positive_distance nanmean: 0.786530 2019-10-09 12:29:52.075: INFO @log_variables: valid negative_distance nanmean: 1.377295 2019-10-09 12:29:52.075: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:29:52.075: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:29:54.325: INFO @metrics_hook: valid matching accuracy: 0.890327996642082, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:29:54.784: INFO @decay_lr : LR updated to `0.00033195384` 2019-10-09 12:29:54.786: INFO @log_profile : T train: 178.280643 2019-10-09 12:29:54.786: INFO @log_profile : T valid: 8.362656 2019-10-09 12:29:54.786: INFO @log_profile : T read data: 1.017497 2019-10-09 12:29:54.786: INFO @log_profile : T hooks: 3.280274 2019-10-09 12:29:54.786: INFO @main_loop : Epoch 220 done 2019-10-09 12:29:54.786: INFO @main_loop : Training epoch 221 2019-10-09 12:33:03.304: INFO @log_variables: train loss mean: 0.238645 2019-10-09 12:33:03.305: INFO @log_variables: train age_loss mean: 3.973124 2019-10-09 12:33:03.305: INFO @log_variables: train gender_loss mean: 0.031946 2019-10-09 12:33:03.305: INFO @log_variables: train matching_loss nanmean: 0.310540 2019-10-09 12:33:03.305: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:33:03.305: INFO @log_variables: train age_mae mean: 4.446941 2019-10-09 12:33:03.305: INFO @log_variables: train gender_accuracy mean: 0.989190 2019-10-09 12:33:03.305: INFO @log_variables: train positive_distance nanmean: 0.738717 2019-10-09 12:33:03.305: INFO @log_variables: train negative_distance nanmean: 1.407414 2019-10-09 12:33:03.305: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:33:03.305: INFO @log_variables: valid loss mean: 0.457670 2019-10-09 12:33:03.305: INFO @log_variables: valid age_loss mean: 6.645537 2019-10-09 12:33:03.305: INFO @log_variables: valid gender_loss mean: 0.258532 2019-10-09 12:33:03.305: INFO @log_variables: valid matching_loss nanmean: 0.495689 2019-10-09 12:33:03.305: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:33:03.305: INFO @log_variables: valid age_mae mean: 7.128705 2019-10-09 12:33:03.305: INFO @log_variables: valid gender_accuracy mean: 0.930684 2019-10-09 12:33:03.305: INFO @log_variables: valid positive_distance nanmean: 0.788407 2019-10-09 12:33:03.306: INFO @log_variables: valid negative_distance nanmean: 1.379340 2019-10-09 12:33:03.306: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:33:03.306: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:33:05.456: INFO @metrics_hook: valid matching accuracy: 0.8882892606583918, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:33:05.901: INFO @decay_lr : LR updated to `0.00033029407` 2019-10-09 12:33:05.903: INFO @log_profile : T train: 178.124225 2019-10-09 12:33:05.903: INFO @log_profile : T valid: 8.323818 2019-10-09 12:33:05.903: INFO @log_profile : T read data: 1.407661 2019-10-09 12:33:05.903: INFO @log_profile : T hooks: 3.175032 2019-10-09 12:33:05.903: INFO @main_loop : Epoch 221 done 2019-10-09 12:33:05.903: INFO @main_loop : Training epoch 222 2019-10-09 12:36:14.836: INFO @log_variables: train loss mean: 0.239118 2019-10-09 12:36:14.837: INFO @log_variables: train age_loss mean: 3.997086 2019-10-09 12:36:14.837: INFO @log_variables: train gender_loss mean: 0.032954 2019-10-09 12:36:14.837: INFO @log_variables: train matching_loss nanmean: 0.308602 2019-10-09 12:36:14.837: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:36:14.837: INFO @log_variables: train age_mae mean: 4.470781 2019-10-09 12:36:14.837: INFO @log_variables: train gender_accuracy mean: 0.988312 2019-10-09 12:36:14.837: INFO @log_variables: train positive_distance nanmean: 0.737806 2019-10-09 12:36:14.837: INFO @log_variables: train negative_distance nanmean: 1.407613 2019-10-09 12:36:14.837: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:36:14.837: INFO @log_variables: valid loss mean: 0.455212 2019-10-09 12:36:14.837: INFO @log_variables: valid age_loss mean: 6.412749 2019-10-09 12:36:14.837: INFO @log_variables: valid gender_loss mean: 0.271334 2019-10-09 12:36:14.837: INFO @log_variables: valid matching_loss nanmean: 0.498547 2019-10-09 12:36:14.837: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:36:14.837: INFO @log_variables: valid age_mae mean: 6.895753 2019-10-09 12:36:14.837: INFO @log_variables: valid gender_accuracy mean: 0.927017 2019-10-09 12:36:14.837: INFO @log_variables: valid positive_distance nanmean: 0.783206 2019-10-09 12:36:14.837: INFO @log_variables: valid negative_distance nanmean: 1.376018 2019-10-09 12:36:14.837: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:36:14.838: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:36:17.349: INFO @metrics_hook: valid matching accuracy: 0.8889488517119386, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:36:17.823: INFO @decay_lr : LR updated to `0.0003286426` 2019-10-09 12:36:17.825: INFO @log_profile : T train: 178.370532 2019-10-09 12:36:17.825: INFO @log_profile : T valid: 8.395019 2019-10-09 12:36:17.825: INFO @log_profile : T read data: 1.499299 2019-10-09 12:36:17.825: INFO @log_profile : T hooks: 3.570704 2019-10-09 12:36:17.825: INFO @main_loop : Epoch 222 done 2019-10-09 12:36:17.825: INFO @main_loop : Training epoch 223 2019-10-09 12:39:26.026: INFO @log_variables: train loss mean: 0.237369 2019-10-09 12:39:26.026: INFO @log_variables: train age_loss mean: 3.950413 2019-10-09 12:39:26.026: INFO @log_variables: train gender_loss mean: 0.033229 2019-10-09 12:39:26.026: INFO @log_variables: train matching_loss nanmean: 0.307574 2019-10-09 12:39:26.026: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:39:26.026: INFO @log_variables: train age_mae mean: 4.423895 2019-10-09 12:39:26.026: INFO @log_variables: train gender_accuracy mean: 0.988658 2019-10-09 12:39:26.026: INFO @log_variables: train positive_distance nanmean: 0.736689 2019-10-09 12:39:26.027: INFO @log_variables: train negative_distance nanmean: 1.407904 2019-10-09 12:39:26.027: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:39:26.027: INFO @log_variables: valid loss mean: 0.457519 2019-10-09 12:39:26.027: INFO @log_variables: valid age_loss mean: 6.535664 2019-10-09 12:39:26.027: INFO @log_variables: valid gender_loss mean: 0.271415 2019-10-09 12:39:26.027: INFO @log_variables: valid matching_loss nanmean: 0.493328 2019-10-09 12:39:26.027: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:39:26.027: INFO @log_variables: valid age_mae mean: 7.019628 2019-10-09 12:39:26.027: INFO @log_variables: valid gender_accuracy mean: 0.928850 2019-10-09 12:39:26.027: INFO @log_variables: valid positive_distance nanmean: 0.782226 2019-10-09 12:39:26.027: INFO @log_variables: valid negative_distance nanmean: 1.377494 2019-10-09 12:39:26.027: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:39:26.027: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:39:28.411: INFO @metrics_hook: valid matching accuracy: 0.8912874018108773, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:39:28.842: INFO @decay_lr : LR updated to `0.0003269994` 2019-10-09 12:39:28.844: INFO @log_profile : T train: 178.214994 2019-10-09 12:39:28.844: INFO @log_profile : T valid: 8.343204 2019-10-09 12:39:28.844: INFO @log_profile : T read data: 0.987167 2019-10-09 12:39:28.844: INFO @log_profile : T hooks: 3.388502 2019-10-09 12:39:28.844: INFO @main_loop : Epoch 223 done 2019-10-09 12:39:28.844: INFO @main_loop : Training epoch 224 2019-10-09 12:42:37.487: INFO @log_variables: train loss mean: 0.236312 2019-10-09 12:42:37.487: INFO @log_variables: train age_loss mean: 3.950536 2019-10-09 12:42:37.487: INFO @log_variables: train gender_loss mean: 0.030060 2019-10-09 12:42:37.487: INFO @log_variables: train matching_loss nanmean: 0.307453 2019-10-09 12:42:37.487: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:42:37.487: INFO @log_variables: train age_mae mean: 4.423790 2019-10-09 12:42:37.487: INFO @log_variables: train gender_accuracy mean: 0.989706 2019-10-09 12:42:37.487: INFO @log_variables: train positive_distance nanmean: 0.735868 2019-10-09 12:42:37.487: INFO @log_variables: train negative_distance nanmean: 1.407579 2019-10-09 12:42:37.487: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:42:37.487: INFO @log_variables: valid loss mean: 0.449648 2019-10-09 12:42:37.488: INFO @log_variables: valid age_loss mean: 6.322967 2019-10-09 12:42:37.488: INFO @log_variables: valid gender_loss mean: 0.269435 2019-10-09 12:42:37.488: INFO @log_variables: valid matching_loss nanmean: 0.492175 2019-10-09 12:42:37.488: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:42:37.488: INFO @log_variables: valid age_mae mean: 6.805232 2019-10-09 12:42:37.488: INFO @log_variables: valid gender_accuracy mean: 0.927372 2019-10-09 12:42:37.488: INFO @log_variables: valid positive_distance nanmean: 0.788757 2019-10-09 12:42:37.488: INFO @log_variables: valid negative_distance nanmean: 1.379288 2019-10-09 12:42:37.488: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:42:37.488: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:42:39.998: INFO @metrics_hook: valid matching accuracy: 0.8916471787491755, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:42:40.446: INFO @decay_lr : LR updated to `0.0003253644` 2019-10-09 12:42:40.448: INFO @log_profile : T train: 178.200748 2019-10-09 12:42:40.448: INFO @log_profile : T valid: 8.378251 2019-10-09 12:42:40.448: INFO @log_profile : T read data: 1.411069 2019-10-09 12:42:40.448: INFO @log_profile : T hooks: 3.526850 2019-10-09 12:42:40.448: INFO @main_loop : Epoch 224 done 2019-10-09 12:42:40.448: INFO @main_loop : Training epoch 225 2019-10-09 12:45:49.079: INFO @log_variables: train loss mean: 0.238039 2019-10-09 12:45:49.079: INFO @log_variables: train age_loss mean: 3.964308 2019-10-09 12:45:49.079: INFO @log_variables: train gender_loss mean: 0.030060 2019-10-09 12:45:49.079: INFO @log_variables: train matching_loss nanmean: 0.311430 2019-10-09 12:45:49.079: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:45:49.079: INFO @log_variables: train age_mae mean: 4.437888 2019-10-09 12:45:49.079: INFO @log_variables: train gender_accuracy mean: 0.989272 2019-10-09 12:45:49.079: INFO @log_variables: train positive_distance nanmean: 0.739376 2019-10-09 12:45:49.079: INFO @log_variables: train negative_distance nanmean: 1.407622 2019-10-09 12:45:49.079: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:45:49.080: INFO @log_variables: valid loss mean: 0.454785 2019-10-09 12:45:49.080: INFO @log_variables: valid age_loss mean: 6.286709 2019-10-09 12:45:49.080: INFO @log_variables: valid gender_loss mean: 0.284711 2019-10-09 12:45:49.080: INFO @log_variables: valid matching_loss nanmean: 0.496451 2019-10-09 12:45:49.080: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:45:49.080: INFO @log_variables: valid age_mae mean: 6.768326 2019-10-09 12:45:49.080: INFO @log_variables: valid gender_accuracy mean: 0.925124 2019-10-09 12:45:49.080: INFO @log_variables: valid positive_distance nanmean: 0.788131 2019-10-09 12:45:49.080: INFO @log_variables: valid negative_distance nanmean: 1.378734 2019-10-09 12:45:49.080: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:45:49.080: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:45:51.223: INFO @metrics_hook: valid matching accuracy: 0.8906877735803802, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:45:51.664: INFO @decay_lr : LR updated to `0.0003237376` 2019-10-09 12:45:51.666: INFO @log_profile : T train: 178.147979 2019-10-09 12:45:51.666: INFO @log_profile : T valid: 8.456539 2019-10-09 12:45:51.666: INFO @log_profile : T read data: 1.375154 2019-10-09 12:45:51.666: INFO @log_profile : T hooks: 3.151608 2019-10-09 12:45:51.666: INFO @main_loop : Epoch 225 done 2019-10-09 12:45:51.666: INFO @main_loop : Training epoch 226 2019-10-09 12:48:59.948: INFO @log_variables: train loss mean: 0.234744 2019-10-09 12:48:59.948: INFO @log_variables: train age_loss mean: 3.920435 2019-10-09 12:48:59.948: INFO @log_variables: train gender_loss mean: 0.031196 2019-10-09 12:48:59.948: INFO @log_variables: train matching_loss nanmean: 0.304467 2019-10-09 12:48:59.948: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:48:59.948: INFO @log_variables: train age_mae mean: 4.393350 2019-10-09 12:48:59.949: INFO @log_variables: train gender_accuracy mean: 0.989225 2019-10-09 12:48:59.949: INFO @log_variables: train positive_distance nanmean: 0.735777 2019-10-09 12:48:59.949: INFO @log_variables: train negative_distance nanmean: 1.407383 2019-10-09 12:48:59.949: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:48:59.949: INFO @log_variables: valid loss mean: 0.459453 2019-10-09 12:48:59.949: INFO @log_variables: valid age_loss mean: 6.510414 2019-10-09 12:48:59.949: INFO @log_variables: valid gender_loss mean: 0.277905 2019-10-09 12:48:59.949: INFO @log_variables: valid matching_loss nanmean: 0.495358 2019-10-09 12:48:59.949: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:48:59.949: INFO @log_variables: valid age_mae mean: 6.993716 2019-10-09 12:48:59.949: INFO @log_variables: valid gender_accuracy mean: 0.927667 2019-10-09 12:48:59.949: INFO @log_variables: valid positive_distance nanmean: 0.792502 2019-10-09 12:48:59.949: INFO @log_variables: valid negative_distance nanmean: 1.379020 2019-10-09 12:48:59.949: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:48:59.949: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:49:02.358: INFO @metrics_hook: valid matching accuracy: 0.8891887030041374, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:49:02.807: INFO @decay_lr : LR updated to `0.0003221189` 2019-10-09 12:49:02.808: INFO @log_profile : T train: 178.246862 2019-10-09 12:49:02.809: INFO @log_profile : T valid: 8.383808 2019-10-09 12:49:02.809: INFO @log_profile : T read data: 0.998104 2019-10-09 12:49:02.809: INFO @log_profile : T hooks: 3.429424 2019-10-09 12:49:02.809: INFO @main_loop : Epoch 226 done 2019-10-09 12:49:02.809: INFO @main_loop : Training epoch 227 2019-10-09 12:52:13.096: INFO @log_variables: train loss mean: 0.234990 2019-10-09 12:52:13.097: INFO @log_variables: train age_loss mean: 3.922170 2019-10-09 12:52:13.097: INFO @log_variables: train gender_loss mean: 0.030427 2019-10-09 12:52:13.097: INFO @log_variables: train matching_loss nanmean: 0.305827 2019-10-09 12:52:13.097: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:52:13.097: INFO @log_variables: train age_mae mean: 4.395491 2019-10-09 12:52:13.097: INFO @log_variables: train gender_accuracy mean: 0.989624 2019-10-09 12:52:13.097: INFO @log_variables: train positive_distance nanmean: 0.735370 2019-10-09 12:52:13.097: INFO @log_variables: train negative_distance nanmean: 1.407332 2019-10-09 12:52:13.097: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:52:13.097: INFO @log_variables: valid loss mean: 0.450033 2019-10-09 12:52:13.097: INFO @log_variables: valid age_loss mean: 6.426280 2019-10-09 12:52:13.097: INFO @log_variables: valid gender_loss mean: 0.258553 2019-10-09 12:52:13.097: INFO @log_variables: valid matching_loss nanmean: 0.493922 2019-10-09 12:52:13.097: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:52:13.097: INFO @log_variables: valid age_mae mean: 6.909906 2019-10-09 12:52:13.097: INFO @log_variables: valid gender_accuracy mean: 0.930151 2019-10-09 12:52:13.098: INFO @log_variables: valid positive_distance nanmean: 0.780846 2019-10-09 12:52:13.098: INFO @log_variables: valid negative_distance nanmean: 1.376417 2019-10-09 12:52:13.098: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:52:13.098: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:52:15.310: INFO @metrics_hook: valid matching accuracy: 0.88864903759669, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:52:15.769: INFO @decay_lr : LR updated to `0.0003205083` 2019-10-09 12:52:15.771: INFO @log_profile : T train: 179.680255 2019-10-09 12:52:15.771: INFO @log_profile : T valid: 8.483008 2019-10-09 12:52:15.771: INFO @log_profile : T read data: 1.427162 2019-10-09 12:52:15.771: INFO @log_profile : T hooks: 3.285198 2019-10-09 12:52:15.771: INFO @main_loop : Epoch 227 done 2019-10-09 12:52:15.771: INFO @main_loop : Training epoch 228 2019-10-09 12:55:24.051: INFO @log_variables: train loss mean: 0.237544 2019-10-09 12:55:24.051: INFO @log_variables: train age_loss mean: 3.973105 2019-10-09 12:55:24.051: INFO @log_variables: train gender_loss mean: 0.030637 2019-10-09 12:55:24.051: INFO @log_variables: train matching_loss nanmean: 0.308437 2019-10-09 12:55:24.051: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:55:24.051: INFO @log_variables: train age_mae mean: 4.446826 2019-10-09 12:55:24.051: INFO @log_variables: train gender_accuracy mean: 0.989000 2019-10-09 12:55:24.052: INFO @log_variables: train positive_distance nanmean: 0.737810 2019-10-09 12:55:24.052: INFO @log_variables: train negative_distance nanmean: 1.407659 2019-10-09 12:55:24.052: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:55:24.052: INFO @log_variables: valid loss mean: 0.457328 2019-10-09 12:55:24.052: INFO @log_variables: valid age_loss mean: 6.415806 2019-10-09 12:55:24.052: INFO @log_variables: valid gender_loss mean: 0.282223 2019-10-09 12:55:24.052: INFO @log_variables: valid matching_loss nanmean: 0.493913 2019-10-09 12:55:24.052: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:55:24.052: INFO @log_variables: valid age_mae mean: 6.899560 2019-10-09 12:55:24.052: INFO @log_variables: valid gender_accuracy mean: 0.927194 2019-10-09 12:55:24.052: INFO @log_variables: valid positive_distance nanmean: 0.786084 2019-10-09 12:55:24.052: INFO @log_variables: valid negative_distance nanmean: 1.376685 2019-10-09 12:55:24.052: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:55:24.052: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:55:26.531: INFO @metrics_hook: valid matching accuracy: 0.8877495952509444, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:55:26.998: INFO @decay_lr : LR updated to `0.00031890575` 2019-10-09 12:55:26.999: INFO @log_profile : T train: 178.234889 2019-10-09 12:55:27.000: INFO @log_profile : T valid: 8.366335 2019-10-09 12:55:27.000: INFO @log_profile : T read data: 1.009153 2019-10-09 12:55:27.000: INFO @log_profile : T hooks: 3.530383 2019-10-09 12:55:27.000: INFO @main_loop : Epoch 228 done 2019-10-09 12:55:27.000: INFO @main_loop : Training epoch 229 2019-10-09 12:58:35.629: INFO @log_variables: train loss mean: 0.237062 2019-10-09 12:58:35.629: INFO @log_variables: train age_loss mean: 3.983115 2019-10-09 12:58:35.629: INFO @log_variables: train gender_loss mean: 0.030768 2019-10-09 12:58:35.629: INFO @log_variables: train matching_loss nanmean: 0.305814 2019-10-09 12:58:35.629: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 12:58:35.629: INFO @log_variables: train age_mae mean: 4.457001 2019-10-09 12:58:35.629: INFO @log_variables: train gender_accuracy mean: 0.988975 2019-10-09 12:58:35.629: INFO @log_variables: train positive_distance nanmean: 0.736628 2019-10-09 12:58:35.630: INFO @log_variables: train negative_distance nanmean: 1.407717 2019-10-09 12:58:35.630: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 12:58:35.630: INFO @log_variables: valid loss mean: 0.448897 2019-10-09 12:58:35.630: INFO @log_variables: valid age_loss mean: 6.369798 2019-10-09 12:58:35.630: INFO @log_variables: valid gender_loss mean: 0.258402 2019-10-09 12:58:35.630: INFO @log_variables: valid matching_loss nanmean: 0.496198 2019-10-09 12:58:35.630: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 12:58:35.630: INFO @log_variables: valid age_mae mean: 6.852701 2019-10-09 12:58:35.630: INFO @log_variables: valid gender_accuracy mean: 0.931571 2019-10-09 12:58:35.630: INFO @log_variables: valid positive_distance nanmean: 0.786063 2019-10-09 12:58:35.630: INFO @log_variables: valid negative_distance nanmean: 1.377965 2019-10-09 12:58:35.630: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 12:58:35.630: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 12:58:38.003: INFO @metrics_hook: valid matching accuracy: 0.8882892606583918, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 12:58:38.460: INFO @decay_lr : LR updated to `0.0003173112` 2019-10-09 12:58:38.462: INFO @log_profile : T train: 178.157522 2019-10-09 12:58:38.462: INFO @log_profile : T valid: 8.342445 2019-10-09 12:58:38.462: INFO @log_profile : T read data: 1.492767 2019-10-09 12:58:38.462: INFO @log_profile : T hooks: 3.383604 2019-10-09 12:58:38.462: INFO @main_loop : Epoch 229 done 2019-10-09 12:58:38.462: INFO @main_loop : Training epoch 230 2019-10-09 13:01:47.418: INFO @log_variables: train loss mean: 0.234686 2019-10-09 13:01:47.418: INFO @log_variables: train age_loss mean: 3.927375 2019-10-09 13:01:47.418: INFO @log_variables: train gender_loss mean: 0.030186 2019-10-09 13:01:47.418: INFO @log_variables: train matching_loss nanmean: 0.304604 2019-10-09 13:01:47.418: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:01:47.418: INFO @log_variables: train age_mae mean: 4.400433 2019-10-09 13:01:47.418: INFO @log_variables: train gender_accuracy mean: 0.989587 2019-10-09 13:01:47.418: INFO @log_variables: train positive_distance nanmean: 0.735216 2019-10-09 13:01:47.418: INFO @log_variables: train negative_distance nanmean: 1.407485 2019-10-09 13:01:47.419: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:01:47.419: INFO @log_variables: valid loss mean: 0.458202 2019-10-09 13:01:47.419: INFO @log_variables: valid age_loss mean: 6.340963 2019-10-09 13:01:47.419: INFO @log_variables: valid gender_loss mean: 0.288735 2019-10-09 13:01:47.419: INFO @log_variables: valid matching_loss nanmean: 0.497595 2019-10-09 13:01:47.419: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:01:47.419: INFO @log_variables: valid age_mae mean: 6.824215 2019-10-09 13:01:47.419: INFO @log_variables: valid gender_accuracy mean: 0.924001 2019-10-09 13:01:47.419: INFO @log_variables: valid positive_distance nanmean: 0.782936 2019-10-09 13:01:47.419: INFO @log_variables: valid negative_distance nanmean: 1.376616 2019-10-09 13:01:47.419: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:01:47.419: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:01:49.553: INFO @metrics_hook: valid matching accuracy: 0.8875097439587456, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:01:50.010: INFO @decay_lr : LR updated to `0.00031572464` 2019-10-09 13:01:50.011: INFO @log_profile : T train: 178.035549 2019-10-09 13:01:50.011: INFO @log_profile : T valid: 8.436831 2019-10-09 13:01:50.011: INFO @log_profile : T read data: 1.830678 2019-10-09 13:01:50.011: INFO @log_profile : T hooks: 3.161078 2019-10-09 13:01:50.011: INFO @main_loop : Epoch 230 done 2019-10-09 13:01:50.012: INFO @main_loop : Training epoch 231 2019-10-09 13:04:58.237: INFO @log_variables: train loss mean: 0.234405 2019-10-09 13:04:58.237: INFO @log_variables: train age_loss mean: 3.918642 2019-10-09 13:04:58.237: INFO @log_variables: train gender_loss mean: 0.029712 2019-10-09 13:04:58.237: INFO @log_variables: train matching_loss nanmean: 0.305080 2019-10-09 13:04:58.237: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:04:58.237: INFO @log_variables: train age_mae mean: 4.391893 2019-10-09 13:04:58.237: INFO @log_variables: train gender_accuracy mean: 0.989715 2019-10-09 13:04:58.238: INFO @log_variables: train positive_distance nanmean: 0.735371 2019-10-09 13:04:58.238: INFO @log_variables: train negative_distance nanmean: 1.407519 2019-10-09 13:04:58.238: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:04:58.238: INFO @log_variables: valid loss mean: 0.455850 2019-10-09 13:04:58.238: INFO @log_variables: valid age_loss mean: 6.478203 2019-10-09 13:04:58.238: INFO @log_variables: valid gender_loss mean: 0.265552 2019-10-09 13:04:58.238: INFO @log_variables: valid matching_loss nanmean: 0.499764 2019-10-09 13:04:58.238: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:04:58.238: INFO @log_variables: valid age_mae mean: 6.961078 2019-10-09 13:04:58.238: INFO @log_variables: valid gender_accuracy mean: 0.927786 2019-10-09 13:04:58.238: INFO @log_variables: valid positive_distance nanmean: 0.787294 2019-10-09 13:04:58.238: INFO @log_variables: valid negative_distance nanmean: 1.378044 2019-10-09 13:04:58.238: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:04:58.239: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:05:00.393: INFO @metrics_hook: valid matching accuracy: 0.8858907477364034, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:05:00.838: INFO @decay_lr : LR updated to `0.00031414602` 2019-10-09 13:05:00.839: INFO @log_profile : T train: 178.102676 2019-10-09 13:05:00.839: INFO @log_profile : T valid: 8.452149 2019-10-09 13:05:00.839: INFO @log_profile : T read data: 0.991489 2019-10-09 13:05:00.839: INFO @log_profile : T hooks: 3.195512 2019-10-09 13:05:00.839: INFO @main_loop : Epoch 231 done 2019-10-09 13:05:00.839: INFO @main_loop : Training epoch 232 2019-10-09 13:08:09.501: INFO @log_variables: train loss mean: 0.233749 2019-10-09 13:08:09.501: INFO @log_variables: train age_loss mean: 3.931340 2019-10-09 13:08:09.501: INFO @log_variables: train gender_loss mean: 0.026463 2019-10-09 13:08:09.501: INFO @log_variables: train matching_loss nanmean: 0.305025 2019-10-09 13:08:09.501: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:08:09.502: INFO @log_variables: train age_mae mean: 4.404407 2019-10-09 13:08:09.502: INFO @log_variables: train gender_accuracy mean: 0.990674 2019-10-09 13:08:09.502: INFO @log_variables: train positive_distance nanmean: 0.735128 2019-10-09 13:08:09.502: INFO @log_variables: train negative_distance nanmean: 1.407622 2019-10-09 13:08:09.502: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:08:09.502: INFO @log_variables: valid loss mean: 0.464799 2019-10-09 13:08:09.502: INFO @log_variables: valid age_loss mean: 6.507268 2019-10-09 13:08:09.502: INFO @log_variables: valid gender_loss mean: 0.289160 2019-10-09 13:08:09.502: INFO @log_variables: valid matching_loss nanmean: 0.500991 2019-10-09 13:08:09.502: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:08:09.502: INFO @log_variables: valid age_mae mean: 6.990953 2019-10-09 13:08:09.502: INFO @log_variables: valid gender_accuracy mean: 0.925302 2019-10-09 13:08:09.502: INFO @log_variables: valid positive_distance nanmean: 0.783301 2019-10-09 13:08:09.502: INFO @log_variables: valid negative_distance nanmean: 1.371480 2019-10-09 13:08:09.503: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:08:09.503: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:08:11.825: INFO @metrics_hook: valid matching accuracy: 0.8869700785512982, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:08:12.269: INFO @decay_lr : LR updated to `0.00031257528` 2019-10-09 13:08:12.270: INFO @log_profile : T train: 178.132943 2019-10-09 13:08:12.270: INFO @log_profile : T valid: 8.486177 2019-10-09 13:08:12.270: INFO @log_profile : T read data: 1.392102 2019-10-09 13:08:12.270: INFO @log_profile : T hooks: 3.334303 2019-10-09 13:08:12.270: INFO @main_loop : Epoch 232 done 2019-10-09 13:08:12.270: INFO @main_loop : Training epoch 233 2019-10-09 13:11:20.940: INFO @log_variables: train loss mean: 0.235196 2019-10-09 13:11:20.940: INFO @log_variables: train age_loss mean: 3.921861 2019-10-09 13:11:20.940: INFO @log_variables: train gender_loss mean: 0.030703 2019-10-09 13:11:20.940: INFO @log_variables: train matching_loss nanmean: 0.306218 2019-10-09 13:11:20.941: INFO @log_variables: train is_face_loss mean: 0.000001 2019-10-09 13:11:20.941: INFO @log_variables: train age_mae mean: 4.395250 2019-10-09 13:11:20.941: INFO @log_variables: train gender_accuracy mean: 0.989553 2019-10-09 13:11:20.941: INFO @log_variables: train positive_distance nanmean: 0.736286 2019-10-09 13:11:20.941: INFO @log_variables: train negative_distance nanmean: 1.407439 2019-10-09 13:11:20.941: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:11:20.941: INFO @log_variables: valid loss mean: 0.456540 2019-10-09 13:11:20.941: INFO @log_variables: valid age_loss mean: 6.295388 2019-10-09 13:11:20.941: INFO @log_variables: valid gender_loss mean: 0.288084 2019-10-09 13:11:20.941: INFO @log_variables: valid matching_loss nanmean: 0.497650 2019-10-09 13:11:20.941: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:11:20.941: INFO @log_variables: valid age_mae mean: 6.777952 2019-10-09 13:11:20.941: INFO @log_variables: valid gender_accuracy mean: 0.926071 2019-10-09 13:11:20.941: INFO @log_variables: valid positive_distance nanmean: 0.786597 2019-10-09 13:11:20.941: INFO @log_variables: valid negative_distance nanmean: 1.377049 2019-10-09 13:11:20.941: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:11:20.941: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:11:23.298: INFO @metrics_hook: valid matching accuracy: 0.8887090004197398, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:11:23.738: INFO @decay_lr : LR updated to `0.0003110124` 2019-10-09 13:11:23.739: INFO @log_profile : T train: 178.308375 2019-10-09 13:11:23.739: INFO @log_profile : T valid: 8.327236 2019-10-09 13:11:23.739: INFO @log_profile : T read data: 1.378031 2019-10-09 13:11:23.739: INFO @log_profile : T hooks: 3.371641 2019-10-09 13:11:23.740: INFO @main_loop : Epoch 233 done 2019-10-09 13:11:23.740: INFO @main_loop : Training epoch 234 2019-10-09 13:14:32.093: INFO @log_variables: train loss mean: 0.233333 2019-10-09 13:14:32.094: INFO @log_variables: train age_loss mean: 3.897513 2019-10-09 13:14:32.094: INFO @log_variables: train gender_loss mean: 0.029961 2019-10-09 13:14:32.094: INFO @log_variables: train matching_loss nanmean: 0.303618 2019-10-09 13:14:32.094: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:14:32.094: INFO @log_variables: train age_mae mean: 4.370775 2019-10-09 13:14:32.094: INFO @log_variables: train gender_accuracy mean: 0.989660 2019-10-09 13:14:32.094: INFO @log_variables: train positive_distance nanmean: 0.735189 2019-10-09 13:14:32.094: INFO @log_variables: train negative_distance nanmean: 1.407326 2019-10-09 13:14:32.094: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:14:32.094: INFO @log_variables: valid loss mean: 0.459980 2019-10-09 13:14:32.094: INFO @log_variables: valid age_loss mean: 6.355193 2019-10-09 13:14:32.094: INFO @log_variables: valid gender_loss mean: 0.291665 2019-10-09 13:14:32.094: INFO @log_variables: valid matching_loss nanmean: 0.498753 2019-10-09 13:14:32.094: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:14:32.094: INFO @log_variables: valid age_mae mean: 6.838533 2019-10-09 13:14:32.094: INFO @log_variables: valid gender_accuracy mean: 0.926958 2019-10-09 13:14:32.094: INFO @log_variables: valid positive_distance nanmean: 0.791043 2019-10-09 13:14:32.094: INFO @log_variables: valid negative_distance nanmean: 1.377887 2019-10-09 13:14:32.095: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:14:32.095: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:14:34.246: INFO @metrics_hook: valid matching accuracy: 0.8882892606583918, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:14:34.713: INFO @decay_lr : LR updated to `0.00030945733` 2019-10-09 13:14:34.714: INFO @log_profile : T train: 178.394462 2019-10-09 13:14:34.714: INFO @log_profile : T valid: 8.283646 2019-10-09 13:14:34.714: INFO @log_profile : T read data: 1.003952 2019-10-09 13:14:34.714: INFO @log_profile : T hooks: 3.207285 2019-10-09 13:14:34.714: INFO @main_loop : Epoch 234 done 2019-10-09 13:14:34.714: INFO @main_loop : Training epoch 235 2019-10-09 13:17:43.312: INFO @log_variables: train loss mean: 0.235886 2019-10-09 13:17:43.312: INFO @log_variables: train age_loss mean: 3.929533 2019-10-09 13:17:43.312: INFO @log_variables: train gender_loss mean: 0.030763 2019-10-09 13:17:43.312: INFO @log_variables: train matching_loss nanmean: 0.307530 2019-10-09 13:17:43.312: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:17:43.312: INFO @log_variables: train age_mae mean: 4.402633 2019-10-09 13:17:43.312: INFO @log_variables: train gender_accuracy mean: 0.989616 2019-10-09 13:17:43.313: INFO @log_variables: train positive_distance nanmean: 0.736932 2019-10-09 13:17:43.313: INFO @log_variables: train negative_distance nanmean: 1.407381 2019-10-09 13:17:43.313: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:17:43.313: INFO @log_variables: valid loss mean: 0.454682 2019-10-09 13:17:43.313: INFO @log_variables: valid age_loss mean: 6.541608 2019-10-09 13:17:43.313: INFO @log_variables: valid gender_loss mean: 0.254249 2019-10-09 13:17:43.313: INFO @log_variables: valid matching_loss nanmean: 0.501105 2019-10-09 13:17:43.313: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:17:43.313: INFO @log_variables: valid age_mae mean: 7.024357 2019-10-09 13:17:43.313: INFO @log_variables: valid gender_accuracy mean: 0.930625 2019-10-09 13:17:43.313: INFO @log_variables: valid positive_distance nanmean: 0.783296 2019-10-09 13:17:43.313: INFO @log_variables: valid negative_distance nanmean: 1.375431 2019-10-09 13:17:43.313: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:17:43.313: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:17:45.118: INFO @metrics_hook: valid matching accuracy: 0.8889488517119386, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:17:45.593: INFO @decay_lr : LR updated to `0.00030791006` 2019-10-09 13:17:45.594: INFO @log_profile : T train: 177.978093 2019-10-09 13:17:45.594: INFO @log_profile : T valid: 8.554771 2019-10-09 13:17:45.594: INFO @log_profile : T read data: 1.399007 2019-10-09 13:17:45.594: INFO @log_profile : T hooks: 2.863528 2019-10-09 13:17:45.594: INFO @main_loop : Epoch 235 done 2019-10-09 13:17:45.594: INFO @main_loop : Training epoch 236 2019-10-09 13:20:53.900: INFO @log_variables: train loss mean: 0.234585 2019-10-09 13:20:53.900: INFO @log_variables: train age_loss mean: 3.936966 2019-10-09 13:20:53.900: INFO @log_variables: train gender_loss mean: 0.029240 2019-10-09 13:20:53.900: INFO @log_variables: train matching_loss nanmean: 0.304277 2019-10-09 13:20:53.900: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:20:53.900: INFO @log_variables: train age_mae mean: 4.410168 2019-10-09 13:20:53.900: INFO @log_variables: train gender_accuracy mean: 0.989824 2019-10-09 13:20:53.900: INFO @log_variables: train positive_distance nanmean: 0.736338 2019-10-09 13:20:53.901: INFO @log_variables: train negative_distance nanmean: 1.407656 2019-10-09 13:20:53.901: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:20:53.901: INFO @log_variables: valid loss mean: 0.460632 2019-10-09 13:20:53.901: INFO @log_variables: valid age_loss mean: 6.490692 2019-10-09 13:20:53.901: INFO @log_variables: valid gender_loss mean: 0.278573 2019-10-09 13:20:53.901: INFO @log_variables: valid matching_loss nanmean: 0.500316 2019-10-09 13:20:53.901: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:20:53.901: INFO @log_variables: valid age_mae mean: 6.973300 2019-10-09 13:20:53.901: INFO @log_variables: valid gender_accuracy mean: 0.927904 2019-10-09 13:20:53.901: INFO @log_variables: valid positive_distance nanmean: 0.783365 2019-10-09 13:20:53.901: INFO @log_variables: valid negative_distance nanmean: 1.374329 2019-10-09 13:20:53.901: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:20:53.901: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:20:56.398: INFO @metrics_hook: valid matching accuracy: 0.8907477364034299, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:20:56.846: INFO @decay_lr : LR updated to `0.0003063705` 2019-10-09 13:20:56.847: INFO @log_profile : T train: 178.234629 2019-10-09 13:20:56.847: INFO @log_profile : T valid: 8.417431 2019-10-09 13:20:56.847: INFO @log_profile : T read data: 0.978966 2019-10-09 13:20:56.847: INFO @log_profile : T hooks: 3.534951 2019-10-09 13:20:56.847: INFO @main_loop : Epoch 236 done 2019-10-09 13:20:56.847: INFO @main_loop : Training epoch 237 2019-10-09 13:24:05.692: INFO @log_variables: train loss mean: 0.234389 2019-10-09 13:24:05.692: INFO @log_variables: train age_loss mean: 3.931976 2019-10-09 13:24:05.692: INFO @log_variables: train gender_loss mean: 0.031198 2019-10-09 13:24:05.692: INFO @log_variables: train matching_loss nanmean: 0.302209 2019-10-09 13:24:05.692: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:24:05.692: INFO @log_variables: train age_mae mean: 4.405471 2019-10-09 13:24:05.692: INFO @log_variables: train gender_accuracy mean: 0.989182 2019-10-09 13:24:05.693: INFO @log_variables: train positive_distance nanmean: 0.733884 2019-10-09 13:24:05.693: INFO @log_variables: train negative_distance nanmean: 1.407571 2019-10-09 13:24:05.693: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:24:05.693: INFO @log_variables: valid loss mean: 0.459515 2019-10-09 13:24:05.693: INFO @log_variables: valid age_loss mean: 6.437478 2019-10-09 13:24:05.693: INFO @log_variables: valid gender_loss mean: 0.283726 2019-10-09 13:24:05.693: INFO @log_variables: valid matching_loss nanmean: 0.497021 2019-10-09 13:24:05.693: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:24:05.693: INFO @log_variables: valid age_mae mean: 6.920996 2019-10-09 13:24:05.693: INFO @log_variables: valid gender_accuracy mean: 0.926780 2019-10-09 13:24:05.693: INFO @log_variables: valid positive_distance nanmean: 0.790044 2019-10-09 13:24:05.693: INFO @log_variables: valid negative_distance nanmean: 1.381809 2019-10-09 13:24:05.693: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:24:05.693: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:24:08.005: INFO @metrics_hook: valid matching accuracy: 0.8878095580739941, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:24:08.467: INFO @decay_lr : LR updated to `0.00030483864` 2019-10-09 13:24:08.468: INFO @log_profile : T train: 178.276924 2019-10-09 13:24:08.468: INFO @log_profile : T valid: 8.463389 2019-10-09 13:24:08.468: INFO @log_profile : T read data: 1.424873 2019-10-09 13:24:08.468: INFO @log_profile : T hooks: 3.370628 2019-10-09 13:24:08.469: INFO @main_loop : Epoch 237 done 2019-10-09 13:24:08.469: INFO @main_loop : Training epoch 238 2019-10-09 13:27:17.298: INFO @log_variables: train loss mean: 0.235390 2019-10-09 13:27:17.298: INFO @log_variables: train age_loss mean: 3.930207 2019-10-09 13:27:17.298: INFO @log_variables: train gender_loss mean: 0.031449 2019-10-09 13:27:17.298: INFO @log_variables: train matching_loss nanmean: 0.305238 2019-10-09 13:27:17.298: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:27:17.299: INFO @log_variables: train age_mae mean: 4.403436 2019-10-09 13:27:17.299: INFO @log_variables: train gender_accuracy mean: 0.989209 2019-10-09 13:27:17.299: INFO @log_variables: train positive_distance nanmean: 0.736533 2019-10-09 13:27:17.299: INFO @log_variables: train negative_distance nanmean: 1.407324 2019-10-09 13:27:17.299: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:27:17.299: INFO @log_variables: valid loss mean: 0.458624 2019-10-09 13:27:17.299: INFO @log_variables: valid age_loss mean: 6.425036 2019-10-09 13:27:17.299: INFO @log_variables: valid gender_loss mean: 0.278901 2019-10-09 13:27:17.299: INFO @log_variables: valid matching_loss nanmean: 0.500331 2019-10-09 13:27:17.299: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:27:17.299: INFO @log_variables: valid age_mae mean: 6.908357 2019-10-09 13:27:17.299: INFO @log_variables: valid gender_accuracy mean: 0.926248 2019-10-09 13:27:17.299: INFO @log_variables: valid positive_distance nanmean: 0.784133 2019-10-09 13:27:17.299: INFO @log_variables: valid negative_distance nanmean: 1.374411 2019-10-09 13:27:17.299: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:27:17.299: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:27:19.591: INFO @metrics_hook: valid matching accuracy: 0.8897283684115849, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:27:20.048: INFO @decay_lr : LR updated to `0.00030331445` 2019-10-09 13:27:20.049: INFO @log_profile : T train: 178.249777 2019-10-09 13:27:20.049: INFO @log_profile : T valid: 8.417380 2019-10-09 13:27:20.049: INFO @log_profile : T read data: 1.493236 2019-10-09 13:27:20.049: INFO @log_profile : T hooks: 3.333792 2019-10-09 13:27:20.049: INFO @main_loop : Epoch 238 done 2019-10-09 13:27:20.049: INFO @main_loop : Training epoch 239 2019-10-09 13:30:28.171: INFO @log_variables: train loss mean: 0.234525 2019-10-09 13:30:28.171: INFO @log_variables: train age_loss mean: 3.931047 2019-10-09 13:30:28.171: INFO @log_variables: train gender_loss mean: 0.028418 2019-10-09 13:30:28.171: INFO @log_variables: train matching_loss nanmean: 0.305506 2019-10-09 13:30:28.171: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:30:28.171: INFO @log_variables: train age_mae mean: 4.404438 2019-10-09 13:30:28.171: INFO @log_variables: train gender_accuracy mean: 0.990185 2019-10-09 13:30:28.171: INFO @log_variables: train positive_distance nanmean: 0.738002 2019-10-09 13:30:28.171: INFO @log_variables: train negative_distance nanmean: 1.407585 2019-10-09 13:30:28.171: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:30:28.171: INFO @log_variables: valid loss mean: 0.460880 2019-10-09 13:30:28.171: INFO @log_variables: valid age_loss mean: 6.430192 2019-10-09 13:30:28.172: INFO @log_variables: valid gender_loss mean: 0.284188 2019-10-09 13:30:28.172: INFO @log_variables: valid matching_loss nanmean: 0.501521 2019-10-09 13:30:28.172: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:30:28.172: INFO @log_variables: valid age_mae mean: 6.912685 2019-10-09 13:30:28.172: INFO @log_variables: valid gender_accuracy mean: 0.925597 2019-10-09 13:30:28.172: INFO @log_variables: valid positive_distance nanmean: 0.782257 2019-10-09 13:30:28.172: INFO @log_variables: valid negative_distance nanmean: 1.375419 2019-10-09 13:30:28.172: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:30:28.172: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:30:30.638: INFO @metrics_hook: valid matching accuracy: 0.8887689632427894, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:30:31.101: INFO @decay_lr : LR updated to `0.00030179787` 2019-10-09 13:30:31.103: INFO @log_profile : T train: 178.082865 2019-10-09 13:30:31.103: INFO @log_profile : T valid: 8.401230 2019-10-09 13:30:31.103: INFO @log_profile : T read data: 0.977662 2019-10-09 13:30:31.103: INFO @log_profile : T hooks: 3.505152 2019-10-09 13:30:31.103: INFO @main_loop : Epoch 239 done 2019-10-09 13:30:31.103: INFO @main_loop : Training epoch 240 2019-10-09 13:33:39.731: INFO @log_variables: train loss mean: 0.233410 2019-10-09 13:33:39.731: INFO @log_variables: train age_loss mean: 3.917531 2019-10-09 13:33:39.731: INFO @log_variables: train gender_loss mean: 0.029109 2019-10-09 13:33:39.731: INFO @log_variables: train matching_loss nanmean: 0.302709 2019-10-09 13:33:39.731: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:33:39.731: INFO @log_variables: train age_mae mean: 4.390638 2019-10-09 13:33:39.731: INFO @log_variables: train gender_accuracy mean: 0.989580 2019-10-09 13:33:39.731: INFO @log_variables: train positive_distance nanmean: 0.734461 2019-10-09 13:33:39.731: INFO @log_variables: train negative_distance nanmean: 1.407496 2019-10-09 13:33:39.731: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:33:39.731: INFO @log_variables: valid loss mean: 0.464391 2019-10-09 13:33:39.731: INFO @log_variables: valid age_loss mean: 6.360872 2019-10-09 13:33:39.732: INFO @log_variables: valid gender_loss mean: 0.306327 2019-10-09 13:33:39.732: INFO @log_variables: valid matching_loss nanmean: 0.497198 2019-10-09 13:33:39.732: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:33:39.732: INFO @log_variables: valid age_mae mean: 6.843789 2019-10-09 13:33:39.732: INFO @log_variables: valid gender_accuracy mean: 0.922699 2019-10-09 13:33:39.732: INFO @log_variables: valid positive_distance nanmean: 0.788432 2019-10-09 13:33:39.732: INFO @log_variables: valid negative_distance nanmean: 1.376477 2019-10-09 13:33:39.732: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:33:39.732: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:33:41.898: INFO @metrics_hook: valid matching accuracy: 0.8900281825268334, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:33:42.369: INFO @decay_lr : LR updated to `0.0003002889` 2019-10-09 13:33:42.370: INFO @log_profile : T train: 178.164606 2019-10-09 13:33:42.370: INFO @log_profile : T valid: 8.320153 2019-10-09 13:33:42.370: INFO @log_profile : T read data: 1.477086 2019-10-09 13:33:42.370: INFO @log_profile : T hooks: 3.218335 2019-10-09 13:33:42.370: INFO @main_loop : Epoch 240 done 2019-10-09 13:33:42.370: INFO @main_loop : Training epoch 241 2019-10-09 13:36:50.919: INFO @log_variables: train loss mean: 0.232743 2019-10-09 13:36:50.919: INFO @log_variables: train age_loss mean: 3.884960 2019-10-09 13:36:50.919: INFO @log_variables: train gender_loss mean: 0.030690 2019-10-09 13:36:50.920: INFO @log_variables: train matching_loss nanmean: 0.302317 2019-10-09 13:36:50.920: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:36:50.920: INFO @log_variables: train age_mae mean: 4.357569 2019-10-09 13:36:50.920: INFO @log_variables: train gender_accuracy mean: 0.989316 2019-10-09 13:36:50.920: INFO @log_variables: train positive_distance nanmean: 0.736115 2019-10-09 13:36:50.920: INFO @log_variables: train negative_distance nanmean: 1.407630 2019-10-09 13:36:50.920: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:36:50.920: INFO @log_variables: valid loss mean: 0.463136 2019-10-09 13:36:50.920: INFO @log_variables: valid age_loss mean: 6.434960 2019-10-09 13:36:50.920: INFO @log_variables: valid gender_loss mean: 0.293192 2019-10-09 13:36:50.920: INFO @log_variables: valid matching_loss nanmean: 0.499032 2019-10-09 13:36:50.920: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:36:50.920: INFO @log_variables: valid age_mae mean: 6.918276 2019-10-09 13:36:50.920: INFO @log_variables: valid gender_accuracy mean: 0.925242 2019-10-09 13:36:50.920: INFO @log_variables: valid positive_distance nanmean: 0.784885 2019-10-09 13:36:50.920: INFO @log_variables: valid negative_distance nanmean: 1.376596 2019-10-09 13:36:50.920: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:36:50.920: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:36:53.462: INFO @metrics_hook: valid matching accuracy: 0.8859507105594532, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:36:53.913: INFO @decay_lr : LR updated to `0.00029878746` 2019-10-09 13:36:53.915: INFO @log_profile : T train: 178.158936 2019-10-09 13:36:53.915: INFO @log_profile : T valid: 8.343203 2019-10-09 13:36:53.915: INFO @log_profile : T read data: 1.380323 2019-10-09 13:36:53.915: INFO @log_profile : T hooks: 3.575736 2019-10-09 13:36:53.915: INFO @main_loop : Epoch 241 done 2019-10-09 13:36:53.915: INFO @main_loop : Training epoch 242 2019-10-09 13:40:02.245: INFO @log_variables: train loss mean: 0.233448 2019-10-09 13:40:02.245: INFO @log_variables: train age_loss mean: 3.925238 2019-10-09 13:40:02.245: INFO @log_variables: train gender_loss mean: 0.028654 2019-10-09 13:40:02.245: INFO @log_variables: train matching_loss nanmean: 0.302510 2019-10-09 13:40:02.245: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:40:02.245: INFO @log_variables: train age_mae mean: 4.398806 2019-10-09 13:40:02.245: INFO @log_variables: train gender_accuracy mean: 0.989870 2019-10-09 13:40:02.245: INFO @log_variables: train positive_distance nanmean: 0.734154 2019-10-09 13:40:02.245: INFO @log_variables: train negative_distance nanmean: 1.407341 2019-10-09 13:40:02.245: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:40:02.245: INFO @log_variables: valid loss mean: 0.460260 2019-10-09 13:40:02.245: INFO @log_variables: valid age_loss mean: 6.453463 2019-10-09 13:40:02.245: INFO @log_variables: valid gender_loss mean: 0.283668 2019-10-09 13:40:02.245: INFO @log_variables: valid matching_loss nanmean: 0.497790 2019-10-09 13:40:02.246: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:40:02.246: INFO @log_variables: valid age_mae mean: 6.937251 2019-10-09 13:40:02.246: INFO @log_variables: valid gender_accuracy mean: 0.926899 2019-10-09 13:40:02.246: INFO @log_variables: valid positive_distance nanmean: 0.783437 2019-10-09 13:40:02.246: INFO @log_variables: valid negative_distance nanmean: 1.376426 2019-10-09 13:40:02.246: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:40:02.246: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:40:05.440: INFO @metrics_hook: valid matching accuracy: 0.8893685914732865, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:40:05.932: INFO @decay_lr : LR updated to `0.00029729353` 2019-10-09 13:40:05.933: INFO @log_profile : T train: 178.131015 2019-10-09 13:40:05.933: INFO @log_profile : T valid: 8.524216 2019-10-09 13:40:05.933: INFO @log_profile : T read data: 1.016432 2019-10-09 13:40:05.933: INFO @log_profile : T hooks: 4.260460 2019-10-09 13:40:05.933: INFO @main_loop : Epoch 242 done 2019-10-09 13:40:05.933: INFO @main_loop : Training epoch 243 2019-10-09 13:43:14.674: INFO @log_variables: train loss mean: 0.233281 2019-10-09 13:43:14.675: INFO @log_variables: train age_loss mean: 3.895528 2019-10-09 13:43:14.675: INFO @log_variables: train gender_loss mean: 0.029126 2019-10-09 13:43:14.675: INFO @log_variables: train matching_loss nanmean: 0.304491 2019-10-09 13:43:14.675: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:43:14.675: INFO @log_variables: train age_mae mean: 4.368497 2019-10-09 13:43:14.675: INFO @log_variables: train gender_accuracy mean: 0.990176 2019-10-09 13:43:14.675: INFO @log_variables: train positive_distance nanmean: 0.736262 2019-10-09 13:43:14.675: INFO @log_variables: train negative_distance nanmean: 1.407680 2019-10-09 13:43:14.675: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:43:14.675: INFO @log_variables: valid loss mean: 0.457165 2019-10-09 13:43:14.675: INFO @log_variables: valid age_loss mean: 6.270287 2019-10-09 13:43:14.675: INFO @log_variables: valid gender_loss mean: 0.293360 2019-10-09 13:43:14.675: INFO @log_variables: valid matching_loss nanmean: 0.496823 2019-10-09 13:43:14.675: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:43:14.675: INFO @log_variables: valid age_mae mean: 6.753320 2019-10-09 13:43:14.675: INFO @log_variables: valid gender_accuracy mean: 0.924178 2019-10-09 13:43:14.675: INFO @log_variables: valid positive_distance nanmean: 0.790474 2019-10-09 13:43:14.675: INFO @log_variables: valid negative_distance nanmean: 1.379383 2019-10-09 13:43:14.675: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:43:14.676: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:43:16.978: INFO @metrics_hook: valid matching accuracy: 0.8882892606583918, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:43:17.428: INFO @decay_lr : LR updated to `0.00029580705` 2019-10-09 13:43:17.429: INFO @log_profile : T train: 178.396324 2019-10-09 13:43:17.429: INFO @log_profile : T valid: 8.281469 2019-10-09 13:43:17.429: INFO @log_profile : T read data: 1.408520 2019-10-09 13:43:17.429: INFO @log_profile : T hooks: 3.321599 2019-10-09 13:43:17.430: INFO @main_loop : Epoch 243 done 2019-10-09 13:43:17.430: INFO @main_loop : Training epoch 244 2019-10-09 13:46:26.000: INFO @log_variables: train loss mean: 0.232167 2019-10-09 13:46:26.000: INFO @log_variables: train age_loss mean: 3.884007 2019-10-09 13:46:26.000: INFO @log_variables: train gender_loss mean: 0.028671 2019-10-09 13:46:26.000: INFO @log_variables: train matching_loss nanmean: 0.302646 2019-10-09 13:46:26.000: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:46:26.000: INFO @log_variables: train age_mae mean: 4.356934 2019-10-09 13:46:26.000: INFO @log_variables: train gender_accuracy mean: 0.989969 2019-10-09 13:46:26.000: INFO @log_variables: train positive_distance nanmean: 0.734824 2019-10-09 13:46:26.001: INFO @log_variables: train negative_distance nanmean: 1.407480 2019-10-09 13:46:26.001: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:46:26.001: INFO @log_variables: valid loss mean: 0.450151 2019-10-09 13:46:26.001: INFO @log_variables: valid age_loss mean: 6.310654 2019-10-09 13:46:26.001: INFO @log_variables: valid gender_loss mean: 0.267918 2019-10-09 13:46:26.001: INFO @log_variables: valid matching_loss nanmean: 0.496486 2019-10-09 13:46:26.001: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:46:26.001: INFO @log_variables: valid age_mae mean: 6.793232 2019-10-09 13:46:26.001: INFO @log_variables: valid gender_accuracy mean: 0.929619 2019-10-09 13:46:26.001: INFO @log_variables: valid positive_distance nanmean: 0.787593 2019-10-09 13:46:26.001: INFO @log_variables: valid negative_distance nanmean: 1.378747 2019-10-09 13:46:26.001: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:46:26.001: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:46:28.423: INFO @metrics_hook: valid matching accuracy: 0.8885291119505906, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:46:28.892: INFO @decay_lr : LR updated to `0.00029432803` 2019-10-09 13:46:28.893: INFO @log_profile : T train: 178.469888 2019-10-09 13:46:28.893: INFO @log_profile : T valid: 8.416564 2019-10-09 13:46:28.893: INFO @log_profile : T read data: 1.030714 2019-10-09 13:46:28.893: INFO @log_profile : T hooks: 3.457897 2019-10-09 13:46:28.893: INFO @main_loop : Epoch 244 done 2019-10-09 13:46:28.893: INFO @main_loop : Training epoch 245 2019-10-09 13:49:37.504: INFO @log_variables: train loss mean: 0.233054 2019-10-09 13:49:37.504: INFO @log_variables: train age_loss mean: 3.900205 2019-10-09 13:49:37.504: INFO @log_variables: train gender_loss mean: 0.028802 2019-10-09 13:49:37.504: INFO @log_variables: train matching_loss nanmean: 0.303645 2019-10-09 13:49:37.504: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:49:37.504: INFO @log_variables: train age_mae mean: 4.373166 2019-10-09 13:49:37.504: INFO @log_variables: train gender_accuracy mean: 0.989905 2019-10-09 13:49:37.504: INFO @log_variables: train positive_distance nanmean: 0.735570 2019-10-09 13:49:37.504: INFO @log_variables: train negative_distance nanmean: 1.407780 2019-10-09 13:49:37.504: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:49:37.505: INFO @log_variables: valid loss mean: 0.449388 2019-10-09 13:49:37.505: INFO @log_variables: valid age_loss mean: 6.317069 2019-10-09 13:49:37.505: INFO @log_variables: valid gender_loss mean: 0.265508 2019-10-09 13:49:37.505: INFO @log_variables: valid matching_loss nanmean: 0.495886 2019-10-09 13:49:37.505: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:49:37.505: INFO @log_variables: valid age_mae mean: 6.800164 2019-10-09 13:49:37.505: INFO @log_variables: valid gender_accuracy mean: 0.930388 2019-10-09 13:49:37.505: INFO @log_variables: valid positive_distance nanmean: 0.788944 2019-10-09 13:49:37.505: INFO @log_variables: valid negative_distance nanmean: 1.379480 2019-10-09 13:49:37.505: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:49:37.505: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:49:39.926: INFO @metrics_hook: valid matching accuracy: 0.8878095580739941, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:49:40.381: INFO @decay_lr : LR updated to `0.0002928564` 2019-10-09 13:49:40.383: INFO @log_profile : T train: 178.066549 2019-10-09 13:49:40.383: INFO @log_profile : T valid: 8.421987 2019-10-09 13:49:40.383: INFO @log_profile : T read data: 1.439235 2019-10-09 13:49:40.383: INFO @log_profile : T hooks: 3.473042 2019-10-09 13:49:40.383: INFO @main_loop : Epoch 245 done 2019-10-09 13:49:40.383: INFO @main_loop : Training epoch 246 2019-10-09 13:52:48.728: INFO @log_variables: train loss mean: 0.233361 2019-10-09 13:52:48.728: INFO @log_variables: train age_loss mean: 3.901256 2019-10-09 13:52:48.728: INFO @log_variables: train gender_loss mean: 0.029855 2019-10-09 13:52:48.728: INFO @log_variables: train matching_loss nanmean: 0.303438 2019-10-09 13:52:48.728: INFO @log_variables: train is_face_loss mean: 0.000000 2019-10-09 13:52:48.729: INFO @log_variables: train age_mae mean: 4.374242 2019-10-09 13:52:48.729: INFO @log_variables: train gender_accuracy mean: 0.989761 2019-10-09 13:52:48.729: INFO @log_variables: train positive_distance nanmean: 0.735237 2019-10-09 13:52:48.729: INFO @log_variables: train negative_distance nanmean: 1.407683 2019-10-09 13:52:48.729: INFO @log_variables: train is_face_accuracy mean: 1.000000 2019-10-09 13:52:48.729: INFO @log_variables: valid loss mean: 0.454846 2019-10-09 13:52:48.729: INFO @log_variables: valid age_loss mean: 6.437414 2019-10-09 13:52:48.729: INFO @log_variables: valid gender_loss mean: 0.268440 2019-10-09 13:52:48.729: INFO @log_variables: valid matching_loss nanmean: 0.497841 2019-10-09 13:52:48.729: INFO @log_variables: valid is_face_loss mean: 0.000000 2019-10-09 13:52:48.729: INFO @log_variables: valid age_mae mean: 6.920754 2019-10-09 13:52:48.729: INFO @log_variables: valid gender_accuracy mean: 0.930092 2019-10-09 13:52:48.729: INFO @log_variables: valid positive_distance nanmean: 0.792688 2019-10-09 13:52:48.729: INFO @log_variables: valid negative_distance nanmean: 1.379016 2019-10-09 13:52:48.729: INFO @log_variables: valid is_face_accuracy mean: 1.000000 2019-10-09 13:52:48.729: INFO @log_dir : Output dir: ./log/RecognitionQuantized_suspicious-leavitt 2019-10-09 13:52:51.230: INFO @metrics_hook: valid matching accuracy: 0.8891887030041374, N=16677, DB_SIZE=16908, N_IDS=1469 2019-10-09 13:52:51.705: INFO @decay_lr : LR updated to `0.00029139212` 2019-10-09 13:52:51.706: INFO @log_profile : T train: 177.990064 2019-10-09 13:52:51.706: INFO @log_profile : T valid: 8.386713 2019-10-09 13:52:51.706: INFO @log_profile : T read data: 1.342168 2019-10-09 13:52:51.707: INFO @log_profile : T hooks: 3.521033 2019-10-09 13:52:51.707: INFO @main_loop : Epoch 246 done 2019-10-09 13:52:51.707: INFO @main_loop : Training epoch 247