{"adam_epsilon": 1e-08, "alphabet": "gene_prot", "append_eos": true, "asl_gamma_neg": 4.0, "asl_gamma_pos": 1.0, "best_metric_type": "f1", "beta1": 0.9, "beta2": 0.99, "buffer_size": 256, "cache_dir": null, "classifier_activate_func": "gelu", "classifier_size": 128, "codes_file": null, "config_path": "../config/lucatriple2/lucatriple2_config.json", "cross_atten": false, "dataset_name": "DeepAbBindv2_genome", "dataset_type": "protein_protein_gene", "delete_old": false, "dev_data_dir": "../dataset/DeepAbBindv2_genome/protein_protein_gene/binary_class/dev/", "device": "cuda", "do_eval": true, "do_lower_case": false, "do_metrics": true, "do_predict": true, "do_train": true, "dropout_prob": 0.1, "early_stop_begin_epoch": null, "early_stop_num_epoch": null, "emb_activate_func": "gelu", "embedding_complete": true, "embedding_complete_seg_overlap": true, "embedding_fixed_len_a_time": 3072, "embedding_input_size": 2560, "embedding_input_size_a": 2560, "embedding_input_size_b": 2560, "embedding_input_size_c": 2560, "eval_all_checkpoints": false, "evaluate_during_training": true, "evaluate_steps": 1000000, "evaluate_strategy": "step", "fc_activate_func": "gelu", "focal_loss_alpha": 0.7, "focal_loss_gamma": 2.0, "focal_loss_reduce": false, "fp16": false, "fp16_embedding": false, "fp16_opt_level": "O1", "fusion_type": "concat", "gradient_accumulation_steps": 1, "hidden_size": 1024, "ignore_index": -100, "input_mode": "triple", "input_type": "matrix", "intermediate_size": 4096, "label_filepath": "../dataset/DeepAbBindv2_genome/protein_protein_gene/binary_class/label.txt", "label_size": 2, "label_type": "DeepAbBindv2", "learning_rate": 0.0002, "llm_dir": "..", "llm_dirpath": "../llm/models/lucaone_virus/v1.0/token_level,span_level,seq_level/lucaone_virus/20240815023346/checkpoint-step3800000", "llm_step": "3800000", "llm_task_level": "token_level,span_level,seq_level", "llm_time_str": "20240815023346", "llm_type": "lucaone_virus", "llm_version": "v1.0", "local_rank": -1, "log_dir": "../logs/DeepAbBindv2_genome/protein_protein_gene/binary_class/lucatriple2/matrix/20250124160316", "logging_steps": 1000, "loss_reduction": "mean", "loss_type": "bce", "lr_decay_rate": 0.9, "lr_update_strategy": "step", "matrix_add_special_token": true, "matrix_dirpath": "/mnt/sanyuan.hy/matrices/DeepAbBindv2/lucaone_virus/v1.0/20240815023346/3800000", "matrix_embedding_exists": true, "matrix_encoder": false, "matrix_encoder_act": false, "matrix_fc_size": "512", "matrix_max_length": 29905, "matrix_max_length_a": null, "matrix_max_length_b": null, "matrix_max_length_c": null, "matrix_pooling_type": "value_attention", "max_grad_norm": 1.0, "max_sentence_length": null, "max_sentences": null, "max_steps": -1, "model_dirpath": null, "model_type": "lucatriple2", "n_gpu": 1, "no_cuda": false, "no_position_embeddings": true, "no_token_embeddings": true, "no_token_type_embeddings": true, "non_ignore": false, "not_append_eos": false, "not_matrix_encoder_shared": false, "not_prepend_bos": false, "not_save_emb_to_disk": false, "not_seq_encoder_shared": false, "num_attention_heads": 0, "num_hidden_layers": 0, "num_train_epochs": 50, "output_dir": "../models/DeepAbBindv2_genome/protein_protein_gene/binary_class/lucatriple2/matrix/20250124160316", "output_mode": "binary_class", "overwrite_cache": false, "overwrite_output_dir": true, "per_gpu_eval_batch_size": 16, "per_gpu_train_batch_size": 16, "pos_weight": 2.0, "position_embedding_type": "absolute", "prepend_bos": true, "save_all": true, "save_steps": 1000000, "seed": 1221, "self_atten": false, "seq_fc_size": "null", "seq_max_length": 29905, "seq_max_length_a": null, "seq_max_length_b": null, "seq_max_length_c": null, "seq_pooling_type": "value_attention", "seq_subword": false, "seq_vocab_path": "gene_prot", "sigmoid": true, "task_level_type": "seq_level", "task_type": "binary_class", "tb_log_dir": "../tb-logs/DeepAbBindv2_genome/protein_protein_gene/binary_class/lucatriple2/matrix/20250124160316", "test_data_dir": "../dataset/DeepAbBindv2_genome/protein_protein_gene/binary_class/test/", "time_str": "20250124160323", "train_data_dir": "../dataset/DeepAbBindv2_genome/protein_protein_gene/binary_class/train/", "trunc_type": "right", "vector_dirpath": "/mnt/sanyuan.hy/vectors/DeepAbBindv2/lucaone_virus/v1.0/20240815023346/3800000", "vector_fc_size": "null", "vocab_size": 39, "warmup_steps": 1000, "weight": null, "weight_decay": 0.01, "worker_num": 1} ################################################## n_gpu: 1 ################################################## Inputs: Input Name List: protein_protein_gene,embedding_matrix_a,embedding_matrix_b,embedding_matrix_c ################################################## Encoder Config: {'llm_type': 'lucaone_virus', 'llm_version': 'v1.0', 'llm_step': '3800000', 'llm_dirpath': '../llm/models/lucaone_virus/v1.0/token_level,span_level,seq_level/lucaone_virus/20240815023346/checkpoint-step3800000', 'input_type': 'matrix', 'trunc_type': 'right', 'seq_max_length': 29905, 'atom_seq_max_length': None, 'vector_dirpath': '/mnt/sanyuan.hy/vectors/DeepAbBindv2/lucaone_virus/v1.0/20240815023346/3800000', 'matrix_dirpath': '/mnt/sanyuan.hy/matrices/DeepAbBindv2/lucaone_virus/v1.0/20240815023346/3800000', 'local_rank': -1, 'max_sentence_length': None, 'max_sentences': None, 'matrix_add_special_token': True, 'embedding_complete': True, 'embedding_complete_seg_overlap': True, 'embedding_fixed_len_a_time': 3072, 'matrix_embedding_exists': True, 'save_emb_to_disk': True, 'fp16_embedding': False} ################################################## Model Config: LucaConfig { "alphabet": "gene_prot", "attention_probs_dropout_prob": 0.1, "classifier_activate_func": "gelu", "classifier_dropout_prob": 0.1, "classifier_size": 128, "cls_token_id": 2, "cross_atten": false, "directionality": "bidi", "emb_activate_func": "gelu", "embedding_input_size": 2560, "embedding_input_size_a": 2560, "embedding_input_size_b": 2560, "embedding_input_size_c": 2560, "fc_activate_func": "gelu", "hidden_act": "gelu", "hidden_dropout_prob": 0.1, "hidden_size": 1024, "ignore_index": -100, "initializer_range": 0.02, "intermediate_size": 4096, "kernel_size": 7, "layer_norm_eps": 1e-12, "layer_norm_type": "post", "loss_reduction": "mean", "matrix_fc_size": [ 512 ], "matrix_max_length": 29905, "matrix_max_length_a": 29905, "matrix_max_length_b": 29905, "matrix_max_length_c": 29905, "matrix_pooling_type": "value_attention", "max_position_embeddings": 29907, "no_position_embeddings": true, "no_token_embeddings": true, "no_token_type_embeddings": true, "num_attention_heads": 4, "num_hidden_layers": 4, "pad_token_id": 0, "pos_weight": 2.0, "position_embedding_type": "absolute", "self_atten": false, "sep_token_id": 3, "seq_fc_size": null, "seq_max_length": 29905, "seq_max_length_a": 29905, "seq_max_length_b": 29905, "seq_max_length_c": 29905, "seq_pooling_type": "value_attention", "token_dropout": null, "transformers_version": "4.41.2", "type_vocab_size": 2, "use_luca_layer_norm_v2": true, "vector_fc_size": null, "vocab_size": 39 } ################################################## Mode Architecture: LucaTriple2( (matrix_pooler_a): GlobalMaskValueAttentionPooling1D (2560 -> 2560) (matrix_pooler_b): GlobalMaskValueAttentionPooling1D (2560 -> 2560) (matrix_pooler_c): GlobalMaskValueAttentionPooling1D (2560 -> 2560) (linear_a): ModuleList( (0): ModuleList( (0): Linear(in_features=2560, out_features=512, bias=True) (1): GELU(approximate='none') ) ) (linear_b): ModuleList( (0): ModuleList( (0): Linear(in_features=2560, out_features=512, bias=True) (1): GELU(approximate='none') ) ) (linear_c): ModuleList( (0): ModuleList( (0): Linear(in_features=2560, out_features=512, bias=True) (1): GELU(approximate='none') ) ) (dropout): Dropout(p=0.1, inplace=False) (hidden_layer): Linear(in_features=1536, out_features=128, bias=True) (hidden_act): GELU(approximate='none') (classifier): Linear(in_features=128, out_features=1, bias=True) (output): Sigmoid() (loss_fct): MaskedBCEWithLogitsLoss( (criterion): BCEWithLogitsLoss() ) ) ################################################## Model parameters: 63112961 ################################################## {"total_num": "60.190000M", "total_size": "240.760000MB", "param_sum": "60.190000M", "param_size": "240.760000MB", "buffer_sum": "0.000000M", "buffer_size": "0.000000MB", "trainable_num": "60.189210M", "trainable_size": "240.756840MB"} ################################################## Train dataset len: 33572, batch size: 16, batch num: 2099 Train dataset t_total: 104950, max_steps: -1 ***** Running training ***** Train Dataset Num examples = 33572 Train Dataset Num Epochs = 50 Logging Steps = 1000 Saving Steps = 1000000 Evaluating Strategy = step Evaluating Steps = 1000000 steps Train Dataset Instantaneous batch size per GPU = 16 Train Dataset Total train batch size (w. parallel, distributed & accumulation) = 16 Train Dataset Gradient Accumulation steps = 1 Train Dataset Total optimization steps = 104950 ################################################## Training, Epoch: 0001, Batch: 001000, Sample Num: 16000, Cur Loss: 1.24686170, Cur Avg Loss: 0.85711599, Log Avg loss: 0.85711599, Global Avg Loss: 0.85711599, Time: 3.0188 Steps: 1000, Updated lr: 0.000200 Training, Epoch: 0001, Batch: 002000, Sample Num: 32000, Cur Loss: 0.72010386, Cur Avg Loss: 0.76033127, Log Avg loss: 0.66354655, Global Avg Loss: 0.76033127, Time: 3.0125 Steps: 2000, Updated lr: 0.000198 ***** Running evaluation checkpoint-2099 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-2099 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6333.998690, Avg time per batch (s): 3.020000 {"eval_avg_loss": 0.657035, "eval_total_loss": 588.046701, "eval_acc": 0.8032, "eval_prec": 0.823977, "eval_recall": 0.771133, "eval_f1": 0.79668, "eval_roc_auc": 0.89191, "eval_pr_auc": 0.883964, "eval_confusion_matrix": {"tn": 5978, "fp": 1179, "fn": 1638, "tp": 5519}, "eval_mcc2": 0.60765, "eval_mcc": 0.60765, "eval_sn": 0.771133, "eval_sp": 0.835266, "update_flag": true, "test_avg_loss": 1.321677, "test_total_loss": 15.86012, "test_acc": 0.666667, "test_prec": 0.519231, "test_recall": 0.435484, "test_f1": 0.473684, "test_roc_auc": 0.600328, "test_pr_auc": 0.423003, "test_confusion_matrix": {"tn": 93, "fp": 25, "fn": 35, "tp": 27}, "test_mcc2": 0.234444, "test_mcc": 0.234444, "test_sn": 0.435484, "test_sp": 0.788136, "lr": 0.0001978855218855219, "cur_epoch_step": 2099, "train_global_avg_loss": 0.7542648582642507, "train_cur_epoch_loss": 1583.2019374966621, "train_cur_epoch_avg_loss": 0.7542648582642507, "train_cur_epoch_time": 6333.998689889908, "train_cur_epoch_avg_time": 3.0176268174797083, "epoch": 1, "step": 2099} ################################################## Training, Epoch: 0002, Batch: 000901, Sample Num: 14416, Cur Loss: 0.44896320, Cur Avg Loss: 0.54425865, Log Avg loss: 0.55291644, Global Avg Loss: 0.69119299, Time: 3.0272 Steps: 3000, Updated lr: 0.000196 Training, Epoch: 0002, Batch: 001901, Sample Num: 30416, Cur Loss: 0.33083260, Cur Avg Loss: 0.52016826, Log Avg loss: 0.49846282, Global Avg Loss: 0.64301045, Time: 3.0477 Steps: 4000, Updated lr: 0.000194 ***** Running evaluation checkpoint-4198 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-4198 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6338.120292, Avg time per batch (s): 3.020000 {"eval_avg_loss": 0.50931, "eval_total_loss": 455.832139, "eval_acc": 0.802222, "eval_prec": 0.727588, "eval_recall": 0.966187, "eval_f1": 0.830082, "eval_roc_auc": 0.932431, "eval_pr_auc": 0.928919, "eval_confusion_matrix": {"tn": 4568, "fp": 2589, "fn": 242, "tp": 6915}, "eval_mcc2": 0.639824, "eval_mcc": 0.639824, "eval_sn": 0.966187, "eval_sp": 0.638256, "update_flag": true, "test_avg_loss": 1.415348, "test_total_loss": 16.984171, "test_acc": 0.572222, "test_prec": 0.421053, "test_recall": 0.645161, "test_f1": 0.509554, "test_roc_auc": 0.638874, "test_pr_auc": 0.503088, "test_confusion_matrix": {"tn": 63, "fp": 55, "fn": 22, "tp": 40}, "test_mcc2": 0.170437, "test_mcc": 0.170437, "test_sn": 0.645161, "test_sp": 0.533898, "lr": 0.00019384704184704187, "cur_epoch_step": 2099, "train_global_avg_loss": 0.6348034273966146, "train_cur_epoch_loss": 1081.702850714326, "train_cur_epoch_avg_loss": 0.5153419965289785, "train_cur_epoch_time": 6338.120292425156, "train_cur_epoch_avg_time": 3.019590420402647, "epoch": 2, "step": 4198} ################################################## Training, Epoch: 0003, Batch: 000802, Sample Num: 12832, Cur Loss: 0.48485768, Cur Avg Loss: 0.36064847, Log Avg loss: 0.38210306, Global Avg Loss: 0.59082897, Time: 3.0284 Steps: 5000, Updated lr: 0.000192 Training, Epoch: 0003, Batch: 001802, Sample Num: 28832, Cur Loss: 0.34032819, Cur Avg Loss: 0.35694838, Log Avg loss: 0.35398090, Global Avg Loss: 0.55135429, Time: 3.0424 Steps: 6000, Updated lr: 0.000190 ***** Running evaluation checkpoint-6297 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-6297 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6354.070842, Avg time per batch (s): 3.030000 {"eval_avg_loss": 0.415442, "eval_total_loss": 371.820656, "eval_acc": 0.875856, "eval_prec": 0.847995, "eval_recall": 0.915887, "eval_f1": 0.880634, "eval_roc_auc": 0.95166, "eval_pr_auc": 0.949755, "eval_confusion_matrix": {"tn": 5982, "fp": 1175, "fn": 602, "tp": 6555}, "eval_mcc2": 0.754132, "eval_mcc": 0.754132, "eval_sn": 0.915887, "eval_sp": 0.835825, "update_flag": true, "test_avg_loss": 1.849054, "test_total_loss": 22.188652, "test_acc": 0.666667, "test_prec": 0.517241, "test_recall": 0.483871, "test_f1": 0.5, "test_roc_auc": 0.672909, "test_pr_auc": 0.500801, "test_confusion_matrix": {"tn": 90, "fp": 28, "fn": 32, "tp": 30}, "test_mcc2": 0.25073, "test_mcc": 0.25073, "test_sn": 0.483871, "test_sp": 0.762712, "lr": 0.00018980856180856183, "cur_epoch_step": 2099, "train_global_avg_loss": 0.5431079947707202, "train_cur_epoch_loss": 755.0462548602372, "train_cur_epoch_avg_loss": 0.35971712951893153, "train_cur_epoch_time": 6354.070841550827, "train_cur_epoch_avg_time": 3.0271895386140195, "epoch": 3, "step": 6297} ################################################## Training, Epoch: 0004, Batch: 000703, Sample Num: 11248, Cur Loss: 0.15241981, Cur Avg Loss: 0.26507877, Log Avg loss: 0.29817565, Global Avg Loss: 0.51518592, Time: 3.0424 Steps: 7000, Updated lr: 0.000188 Training, Epoch: 0004, Batch: 001703, Sample Num: 27248, Cur Loss: 0.11439108, Cur Avg Loss: 0.26282532, Log Avg loss: 0.26124114, Global Avg Loss: 0.48344282, Time: 3.0267 Steps: 8000, Updated lr: 0.000187 ***** Running evaluation checkpoint-8396 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-8396 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6365.239350, Avg time per batch (s): 3.030000 {"eval_avg_loss": 0.439872, "eval_total_loss": 393.685036, "eval_acc": 0.873411, "eval_prec": 0.824529, "eval_recall": 0.948722, "eval_f1": 0.882277, "eval_roc_auc": 0.958257, "eval_pr_auc": 0.956231, "eval_confusion_matrix": {"tn": 5712, "fp": 1445, "fn": 367, "tp": 6790}, "eval_mcc2": 0.75544, "eval_mcc": 0.75544, "eval_sn": 0.948722, "eval_sp": 0.7981, "update_flag": true, "test_avg_loss": 2.260395, "test_total_loss": 27.124739, "test_acc": 0.655556, "test_prec": 0.5, "test_recall": 0.580645, "test_f1": 0.537313, "test_roc_auc": 0.683844, "test_pr_auc": 0.54596, "test_confusion_matrix": {"tn": 82, "fp": 36, "fn": 26, "tp": 36}, "test_mcc2": 0.267286, "test_mcc": 0.267286, "test_sn": 0.580645, "test_sp": 0.694915, "lr": 0.00018577008177008176, "cur_epoch_step": 2099, "train_global_avg_loss": 0.4729868067480493, "train_cur_epoch_loss": 551.2461863853969, "train_cur_epoch_avg_loss": 0.2626232426800366, "train_cur_epoch_time": 6365.239350318909, "train_cur_epoch_avg_time": 3.0325104098708473, "epoch": 4, "step": 8396} ################################################## Training, Epoch: 0005, Batch: 000604, Sample Num: 9664, Cur Loss: 0.10056508, Cur Avg Loss: 0.18909297, Log Avg loss: 0.21786683, Global Avg Loss: 0.45393438, Time: 3.0230 Steps: 9000, Updated lr: 0.000185 Training, Epoch: 0005, Batch: 001604, Sample Num: 25664, Cur Loss: 0.05706108, Cur Avg Loss: 0.19034688, Log Avg loss: 0.19110425, Global Avg Loss: 0.42765136, Time: 3.0127 Steps: 10000, Updated lr: 0.000183 ***** Running evaluation checkpoint-10495 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-10495 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6347.940577, Avg time per batch (s): 3.020000 {"eval_avg_loss": 0.501876, "eval_total_loss": 449.179228, "eval_acc": 0.863979, "eval_prec": 0.804892, "eval_recall": 0.960877, "eval_f1": 0.875995, "eval_roc_auc": 0.958806, "eval_pr_auc": 0.956885, "eval_confusion_matrix": {"tn": 5490, "fp": 1667, "fn": 280, "tp": 6877}, "eval_mcc2": 0.742026, "eval_mcc": 0.742026, "eval_sn": 0.960877, "eval_sp": 0.767081, "update_flag": false, "test_avg_loss": 2.336329, "test_total_loss": 28.035951, "test_acc": 0.655556, "test_prec": 0.5, "test_recall": 0.612903, "test_f1": 0.550725, "test_roc_auc": 0.693685, "test_pr_auc": 0.567054, "test_confusion_matrix": {"tn": 80, "fp": 38, "fn": 24, "tp": 38}, "test_mcc2": 0.279841, "test_mcc": 0.279841, "test_sn": 0.612903, "test_sp": 0.677966, "lr": 0.00018173160173160172, "cur_epoch_step": 2099, "train_global_avg_loss": 0.4173066627321185, "train_cur_epoch_loss": 408.4361959169619, "train_cur_epoch_avg_loss": 0.19458608666839536, "train_cur_epoch_time": 6347.940577030182, "train_cur_epoch_avg_time": 3.0242689742878426, "epoch": 5, "step": 10495} ################################################## Training, Epoch: 0006, Batch: 000505, Sample Num: 8080, Cur Loss: 0.25658602, Cur Avg Loss: 0.14089316, Log Avg loss: 0.17427084, Global Avg Loss: 0.40461677, Time: 3.0172 Steps: 11000, Updated lr: 0.000181 Training, Epoch: 0006, Batch: 001505, Sample Num: 24080, Cur Loss: 0.28952017, Cur Avg Loss: 0.15249980, Log Avg loss: 0.15836115, Global Avg Loss: 0.38409547, Time: 3.0268 Steps: 12000, Updated lr: 0.000179 ***** Running evaluation checkpoint-12594 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-12594 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6353.689340, Avg time per batch (s): 3.030000 {"eval_avg_loss": 0.534453, "eval_total_loss": 478.335394, "eval_acc": 0.889479, "eval_prec": 0.857051, "eval_recall": 0.934889, "eval_f1": 0.89428, "eval_roc_auc": 0.959469, "eval_pr_auc": 0.958299, "eval_confusion_matrix": {"tn": 6041, "fp": 1116, "fn": 466, "tp": 6691}, "eval_mcc2": 0.78219, "eval_mcc": 0.78219, "eval_sn": 0.934889, "eval_sp": 0.844069, "update_flag": true, "test_avg_loss": 2.948665, "test_total_loss": 35.383977, "test_acc": 0.705556, "test_prec": 0.571429, "test_recall": 0.580645, "test_f1": 0.576, "test_roc_auc": 0.677283, "test_pr_auc": 0.554052, "test_confusion_matrix": {"tn": 91, "fp": 27, "fn": 26, "tp": 36}, "test_mcc2": 0.350517, "test_mcc": 0.350517, "test_sn": 0.580645, "test_sp": 0.771186, "lr": 0.0001776931216931217, "cur_epoch_step": 2099, "train_global_avg_loss": 0.37366548785291137, "train_cur_epoch_loss": 326.3097286459815, "train_cur_epoch_avg_loss": 0.15545961345687542, "train_cur_epoch_time": 6353.689339637756, "train_cur_epoch_avg_time": 3.0270077844867824, "epoch": 6, "step": 12594} ################################################## Training, Epoch: 0007, Batch: 000406, Sample Num: 6496, Cur Loss: 0.01214808, Cur Avg Loss: 0.11769259, Log Avg loss: 0.14458072, Global Avg Loss: 0.36567126, Time: 3.0134 Steps: 13000, Updated lr: 0.000177 Training, Epoch: 0007, Batch: 001406, Sample Num: 22496, Cur Loss: 0.01397178, Cur Avg Loss: 0.11927828, Log Avg loss: 0.11992206, Global Avg Loss: 0.34811774, Time: 3.0192 Steps: 14000, Updated lr: 0.000175 ***** Running evaluation checkpoint-14693 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-14693 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6349.999820, Avg time per batch (s): 3.030000 {"eval_avg_loss": 0.729703, "eval_total_loss": 653.083962, "eval_acc": 0.902962, "eval_prec": 0.901671, "eval_recall": 0.904569, "eval_f1": 0.903118, "eval_roc_auc": 0.960626, "eval_pr_auc": 0.961487, "eval_confusion_matrix": {"tn": 6451, "fp": 706, "fn": 683, "tp": 6474}, "eval_mcc2": 0.805928, "eval_mcc": 0.805928, "eval_sn": 0.904569, "eval_sp": 0.901355, "update_flag": true, "test_avg_loss": 3.907293, "test_total_loss": 46.887516, "test_acc": 0.75, "test_prec": 0.666667, "test_recall": 0.548387, "test_f1": 0.60177, "test_roc_auc": 0.698742, "test_pr_auc": 0.565617, "test_confusion_matrix": {"tn": 101, "fp": 17, "fn": 28, "tp": 34}, "test_mcc2": 0.426365, "test_mcc": 0.426365, "test_sn": 0.548387, "test_sp": 0.855932, "lr": 0.00017365464165464167, "cur_epoch_step": 2099, "train_global_avg_loss": 0.3378914564410511, "train_cur_epoch_loss": 258.6960154687986, "train_cur_epoch_avg_loss": 0.12324726796988975, "train_cur_epoch_time": 6349.999820232391, "train_cur_epoch_avg_time": 3.0252500334599293, "epoch": 7, "step": 14693} ################################################## Training, Epoch: 0008, Batch: 000307, Sample Num: 4912, Cur Loss: 0.07990419, Cur Avg Loss: 0.07915328, Log Avg loss: 0.11529082, Global Avg Loss: 0.33259595, Time: 3.0268 Steps: 15000, Updated lr: 0.000173 Training, Epoch: 0008, Batch: 001307, Sample Num: 20912, Cur Loss: 0.06177119, Cur Avg Loss: 0.10143342, Log Avg loss: 0.10827342, Global Avg Loss: 0.31857579, Time: 3.0321 Steps: 16000, Updated lr: 0.000171 ***** Running evaluation checkpoint-16792 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-16792 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6347.890741, Avg time per batch (s): 3.020000 {"eval_avg_loss": 0.657855, "eval_total_loss": 588.780213, "eval_acc": 0.891505, "eval_prec": 0.857854, "eval_recall": 0.938522, "eval_f1": 0.896377, "eval_roc_auc": 0.959929, "eval_pr_auc": 0.959157, "eval_confusion_matrix": {"tn": 6044, "fp": 1113, "fn": 440, "tp": 6717}, "eval_mcc2": 0.786495, "eval_mcc": 0.786495, "eval_sn": 0.938522, "eval_sp": 0.844488, "update_flag": false, "test_avg_loss": 3.8935, "test_total_loss": 46.721997, "test_acc": 0.688889, "test_prec": 0.548387, "test_recall": 0.548387, "test_f1": 0.548387, "test_roc_auc": 0.680563, "test_pr_auc": 0.555801, "test_confusion_matrix": {"tn": 90, "fp": 28, "fn": 28, "tp": 34}, "test_mcc2": 0.311099, "test_mcc": 0.311099, "test_sn": 0.548387, "test_sp": 0.762712, "lr": 0.00016961616161616163, "cur_epoch_step": 2099, "train_global_avg_loss": 0.3092743751959825, "train_cur_epoch_loss": 228.69613880257384, "train_cur_epoch_avg_loss": 0.10895480648050207, "train_cur_epoch_time": 6347.890741348267, "train_cur_epoch_avg_time": 3.0242452317047483, "epoch": 8, "step": 16792} ################################################## Training, Epoch: 0009, Batch: 000208, Sample Num: 3328, Cur Loss: 0.05298880, Cur Avg Loss: 0.07708193, Log Avg loss: 0.11215570, Global Avg Loss: 0.30643343, Time: 3.0184 Steps: 17000, Updated lr: 0.000169 Training, Epoch: 0009, Batch: 001208, Sample Num: 19328, Cur Loss: 0.00733288, Cur Avg Loss: 0.09283313, Log Avg loss: 0.09610938, Global Avg Loss: 0.29474876, Time: 3.0378 Steps: 18000, Updated lr: 0.000167 ***** Running evaluation checkpoint-18891 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-18891 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6371.840143, Avg time per batch (s): 3.040000 {"eval_avg_loss": 0.80098, "eval_total_loss": 716.876692, "eval_acc": 0.9038, "eval_prec": 0.889803, "eval_recall": 0.921755, "eval_f1": 0.905497, "eval_roc_auc": 0.960462, "eval_pr_auc": 0.96063, "eval_confusion_matrix": {"tn": 6340, "fp": 817, "fn": 560, "tp": 6597}, "eval_mcc2": 0.808122, "eval_mcc": 0.808122, "eval_sn": 0.921755, "eval_sp": 0.885846, "update_flag": true, "test_avg_loss": 4.796487, "test_total_loss": 57.557843, "test_acc": 0.705556, "test_prec": 0.581818, "test_recall": 0.516129, "test_f1": 0.547009, "test_roc_auc": 0.68193, "test_pr_auc": 0.581269, "test_confusion_matrix": {"tn": 95, "fp": 23, "fn": 30, "tp": 32}, "test_mcc2": 0.331356, "test_mcc": 0.331356, "test_sn": 0.516129, "test_sp": 0.805085, "lr": 0.00016557768157768159, "cur_epoch_step": 2099, "train_global_avg_loss": 0.2857029296468256, "train_cur_epoch_loss": 203.87873566724375, "train_cur_epoch_avg_loss": 0.09713136525357015, "train_cur_epoch_time": 6371.840143442154, "train_cur_epoch_avg_time": 3.0356551421830176, "epoch": 9, "step": 18891} ################################################## Training, Epoch: 0010, Batch: 000109, Sample Num: 1744, Cur Loss: 0.00196799, Cur Avg Loss: 0.05834597, Log Avg loss: 0.09809603, Global Avg Loss: 0.28439862, Time: 3.0407 Steps: 19000, Updated lr: 0.000165 Training, Epoch: 0010, Batch: 001109, Sample Num: 17744, Cur Loss: 0.00134742, Cur Avg Loss: 0.07489911, Log Avg loss: 0.07670341, Global Avg Loss: 0.27401386, Time: 3.0322 Steps: 20000, Updated lr: 0.000163 ***** Running evaluation checkpoint-20990 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-20990 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6354.976831, Avg time per batch (s): 3.030000 {"eval_avg_loss": 0.830919, "eval_total_loss": 743.672315, "eval_acc": 0.894579, "eval_prec": 0.865708, "eval_recall": 0.934051, "eval_f1": 0.898582, "eval_roc_auc": 0.955602, "eval_pr_auc": 0.953756, "eval_confusion_matrix": {"tn": 6120, "fp": 1037, "fn": 472, "tp": 6685}, "eval_mcc2": 0.791628, "eval_mcc": 0.791628, "eval_sn": 0.934051, "eval_sp": 0.855107, "update_flag": false, "test_avg_loss": 4.510451, "test_total_loss": 54.125413, "test_acc": 0.727778, "test_prec": 0.610169, "test_recall": 0.580645, "test_f1": 0.595041, "test_roc_auc": 0.713231, "test_pr_auc": 0.627438, "test_confusion_matrix": {"tn": 95, "fp": 23, "fn": 26, "tp": 36}, "test_mcc2": 0.390482, "test_mcc": 0.390482, "test_sn": 0.580645, "test_sp": 0.805085, "lr": 0.00016153920153920155, "cur_epoch_step": 2099, "train_global_avg_loss": 0.26539751390676053, "train_cur_epoch_loss": 173.47977294472184, "train_cur_epoch_avg_loss": 0.08264877224617524, "train_cur_epoch_time": 6354.97683095932, "train_cur_epoch_avg_time": 3.0276211676795235, "epoch": 10, "step": 20990} ################################################## Training, Epoch: 0011, Batch: 000010, Sample Num: 160, Cur Loss: 0.05762073, Cur Avg Loss: 0.04775681, Log Avg loss: 0.09089422, Global Avg Loss: 0.26529388, Time: 3.0418 Steps: 21000, Updated lr: 0.000162 Training, Epoch: 0011, Batch: 001010, Sample Num: 16160, Cur Loss: 0.02258909, Cur Avg Loss: 0.06463661, Log Avg loss: 0.06480540, Global Avg Loss: 0.25618076, Time: 3.0303 Steps: 22000, Updated lr: 0.000160 Training, Epoch: 0011, Batch: 002010, Sample Num: 32160, Cur Loss: 0.00309792, Cur Avg Loss: 0.07280082, Log Avg loss: 0.08104667, Global Avg Loss: 0.24856624, Time: 3.0464 Steps: 23000, Updated lr: 0.000158 ***** Running evaluation checkpoint-23089 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-23089 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6379.303071, Avg time per batch (s): 3.040000 {"eval_avg_loss": 0.859438, "eval_total_loss": 769.197283, "eval_acc": 0.905407, "eval_prec": 0.891724, "eval_recall": 0.922873, "eval_f1": 0.907031, "eval_roc_auc": 0.960278, "eval_pr_auc": 0.957775, "eval_confusion_matrix": {"tn": 6355, "fp": 802, "fn": 552, "tp": 6605}, "eval_mcc2": 0.81131, "eval_mcc": 0.81131, "eval_sn": 0.922873, "eval_sp": 0.887942, "update_flag": true, "test_avg_loss": 5.586959, "test_total_loss": 67.043504, "test_acc": 0.727778, "test_prec": 0.622642, "test_recall": 0.532258, "test_f1": 0.573913, "test_roc_auc": 0.681247, "test_pr_auc": 0.60678, "test_confusion_matrix": {"tn": 98, "fp": 20, "fn": 29, "tp": 33}, "test_mcc2": 0.378202, "test_mcc": 0.378202, "test_sn": 0.532258, "test_sp": 0.830508, "lr": 0.0001575007215007215, "cur_epoch_step": 2099, "train_global_avg_loss": 0.2479373577593096, "train_cur_epoch_loss": 153.93183640179632, "train_cur_epoch_avg_loss": 0.07333579628480054, "train_cur_epoch_time": 6379.303070783615, "train_cur_epoch_avg_time": 3.039210610187525, "epoch": 11, "step": 23089} ################################################## Training, Epoch: 0012, Batch: 000911, Sample Num: 14576, Cur Loss: 0.07947645, Cur Avg Loss: 0.06014046, Log Avg loss: 0.06239016, Global Avg Loss: 0.24080890, Time: 3.0398 Steps: 24000, Updated lr: 0.000156 Training, Epoch: 0012, Batch: 001911, Sample Num: 30576, Cur Loss: 0.00789209, Cur Avg Loss: 0.06578093, Log Avg loss: 0.07091939, Global Avg Loss: 0.23401332, Time: 3.0515 Steps: 25000, Updated lr: 0.000154 ***** Running evaluation checkpoint-25188 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-25188 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6357.599402, Avg time per batch (s): 3.030000 {"eval_avg_loss": 1.010724, "eval_total_loss": 904.597919, "eval_acc": 0.907783, "eval_prec": 0.899521, "eval_recall": 0.918122, "eval_f1": 0.908726, "eval_roc_auc": 0.962297, "eval_pr_auc": 0.961077, "eval_confusion_matrix": {"tn": 6423, "fp": 734, "fn": 586, "tp": 6571}, "eval_mcc2": 0.81574, "eval_mcc": 0.81574, "eval_sn": 0.918122, "eval_sp": 0.897443, "update_flag": true, "test_avg_loss": 7.11685, "test_total_loss": 85.402206, "test_acc": 0.738889, "test_prec": 0.659574, "test_recall": 0.5, "test_f1": 0.568807, "test_roc_auc": 0.680563, "test_pr_auc": 0.595439, "test_confusion_matrix": {"tn": 102, "fp": 16, "fn": 31, "tp": 31}, "test_mcc2": 0.394229, "test_mcc": 0.394229, "test_sn": 0.5, "test_sp": 0.864407, "lr": 0.00015346224146224147, "cur_epoch_step": 2099, "train_global_avg_loss": 0.23285044400240212, "train_cur_epoch_loss": 140.41133022780454, "train_cur_epoch_avg_loss": 0.0668943926764195, "train_cur_epoch_time": 6357.599401712418, "train_cur_epoch_avg_time": 3.0288706058658494, "epoch": 12, "step": 25188} ################################################## Training, Epoch: 0013, Batch: 000812, Sample Num: 12992, Cur Loss: 0.00378860, Cur Avg Loss: 0.04991583, Log Avg loss: 0.05523564, Global Avg Loss: 0.22713726, Time: 3.0209 Steps: 26000, Updated lr: 0.000152 Training, Epoch: 0013, Batch: 001812, Sample Num: 28992, Cur Loss: 0.01363603, Cur Avg Loss: 0.06195593, Log Avg loss: 0.07173249, Global Avg Loss: 0.22138152, Time: 3.0462 Steps: 27000, Updated lr: 0.000150 ***** Running evaluation checkpoint-27287 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-27287 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6367.389221, Avg time per batch (s): 3.030000 {"eval_avg_loss": 1.019389, "eval_total_loss": 912.352755, "eval_acc": 0.896046, "eval_prec": 0.86826, "eval_recall": 0.933771, "eval_f1": 0.899825, "eval_roc_auc": 0.955833, "eval_pr_auc": 0.953994, "eval_confusion_matrix": {"tn": 6143, "fp": 1014, "fn": 474, "tp": 6683}, "eval_mcc2": 0.794356, "eval_mcc": 0.794356, "eval_sn": 0.933771, "eval_sp": 0.858321, "update_flag": false, "test_avg_loss": 5.47615, "test_total_loss": 65.713794, "test_acc": 0.694444, "test_prec": 0.559322, "test_recall": 0.532258, "test_f1": 0.545455, "test_roc_auc": 0.70339, "test_pr_auc": 0.621538, "test_confusion_matrix": {"tn": 92, "fp": 26, "fn": 29, "tp": 33}, "test_mcc2": 0.315762, "test_mcc": 0.315762, "test_sn": 0.532258, "test_sp": 0.779661, "lr": 0.00014942376142376143, "cur_epoch_step": 2099, "train_global_avg_loss": 0.2196915059674615, "train_cur_epoch_loss": 129.68513980161708, "train_cur_epoch_avg_loss": 0.06178424954817393, "train_cur_epoch_time": 6367.389220952988, "train_cur_epoch_avg_time": 3.0335346455231003, "epoch": 13, "step": 27287} ################################################## Training, Epoch: 0014, Batch: 000713, Sample Num: 11408, Cur Loss: 0.20919219, Cur Avg Loss: 0.05170637, Log Avg loss: 0.05428763, Global Avg Loss: 0.21541388, Time: 3.0283 Steps: 28000, Updated lr: 0.000148 Training, Epoch: 0014, Batch: 001713, Sample Num: 27408, Cur Loss: 0.46864098, Cur Avg Loss: 0.05721260, Log Avg loss: 0.06113855, Global Avg Loss: 0.21009405, Time: 3.0256 Steps: 29000, Updated lr: 0.000146 ***** Running evaluation checkpoint-29386 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-29386 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6360.300734, Avg time per batch (s): 3.030000 {"eval_avg_loss": 1.070177, "eval_total_loss": 957.808641, "eval_acc": 0.907852, "eval_prec": 0.899863, "eval_recall": 0.917843, "eval_f1": 0.908764, "eval_roc_auc": 0.9606, "eval_pr_auc": 0.958819, "eval_confusion_matrix": {"tn": 6426, "fp": 731, "fn": 588, "tp": 6569}, "eval_mcc2": 0.815868, "eval_mcc": 0.815868, "eval_sn": 0.917843, "eval_sp": 0.897862, "update_flag": true, "test_avg_loss": 7.131944, "test_total_loss": 85.583333, "test_acc": 0.722222, "test_prec": 0.630435, "test_recall": 0.467742, "test_f1": 0.537037, "test_roc_auc": 0.685621, "test_pr_auc": 0.572496, "test_confusion_matrix": {"tn": 101, "fp": 17, "fn": 33, "tp": 29}, "test_mcc2": 0.352625, "test_mcc": 0.352625, "test_sn": 0.467742, "test_sp": 0.855932, "lr": 0.0001453852813852814, "cur_epoch_step": 2099, "train_global_avg_loss": 0.20819650039405774, "train_cur_epoch_loss": 123.34023724565878, "train_cur_epoch_avg_loss": 0.058761427939808854, "train_cur_epoch_time": 6360.300733566284, "train_cur_epoch_avg_time": 3.030157567206424, "epoch": 14, "step": 29386} ################################################## Training, Epoch: 0015, Batch: 000614, Sample Num: 9824, Cur Loss: 0.00087065, Cur Avg Loss: 0.03996605, Log Avg loss: 0.04987420, Global Avg Loss: 0.20475338, Time: 3.0524 Steps: 30000, Updated lr: 0.000144 Training, Epoch: 0015, Batch: 001614, Sample Num: 25824, Cur Loss: 0.00507790, Cur Avg Loss: 0.05010989, Log Avg loss: 0.05633820, Global Avg Loss: 0.19996580, Time: 3.0485 Steps: 31000, Updated lr: 0.000142 ***** Running evaluation checkpoint-31485 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-31485 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6353.269949, Avg time per batch (s): 3.030000 {"eval_avg_loss": 1.122643, "eval_total_loss": 1004.765806, "eval_acc": 0.897932, "eval_prec": 0.876023, "eval_recall": 0.927064, "eval_f1": 0.900821, "eval_roc_auc": 0.957291, "eval_pr_auc": 0.956893, "eval_confusion_matrix": {"tn": 6218, "fp": 939, "fn": 522, "tp": 6635}, "eval_mcc2": 0.797219, "eval_mcc": 0.797219, "eval_sn": 0.927064, "eval_sp": 0.8688, "update_flag": false, "test_avg_loss": 6.38176, "test_total_loss": 76.581114, "test_acc": 0.694444, "test_prec": 0.557377, "test_recall": 0.548387, "test_f1": 0.552846, "test_roc_auc": 0.684322, "test_pr_auc": 0.607555, "test_confusion_matrix": {"tn": 91, "fp": 27, "fn": 28, "tp": 34}, "test_mcc2": 0.320826, "test_mcc": 0.320826, "test_sn": 0.548387, "test_sp": 0.771186, "lr": 0.00014134680134680137, "cur_epoch_step": 2099, "train_global_avg_loss": 0.19747347671204213, "train_cur_epoch_loss": 99.39005369886766, "train_cur_epoch_avg_loss": 0.047351145163824514, "train_cur_epoch_time": 6353.269948720932, "train_cur_epoch_avg_time": 3.0268079793811014, "epoch": 15, "step": 31485} ################################################## Training, Epoch: 0016, Batch: 000515, Sample Num: 8240, Cur Loss: 0.05109704, Cur Avg Loss: 0.04232475, Log Avg loss: 0.04030994, Global Avg Loss: 0.19497655, Time: 3.0198 Steps: 32000, Updated lr: 0.000140 Training, Epoch: 0016, Batch: 001515, Sample Num: 24240, Cur Loss: 0.00002336, Cur Avg Loss: 0.04670257, Log Avg loss: 0.04895714, Global Avg Loss: 0.19055172, Time: 3.0486 Steps: 33000, Updated lr: 0.000138 ***** Running evaluation checkpoint-33584 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-33584 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6352.782499, Avg time per batch (s): 3.030000 {"eval_avg_loss": 1.169405, "eval_total_loss": 1046.617177, "eval_acc": 0.905826, "eval_prec": 0.890652, "eval_recall": 0.925248, "eval_f1": 0.907621, "eval_roc_auc": 0.958008, "eval_pr_auc": 0.95401, "eval_confusion_matrix": {"tn": 6344, "fp": 813, "fn": 535, "tp": 6622}, "eval_mcc2": 0.812266, "eval_mcc": 0.812266, "eval_sn": 0.925248, "eval_sp": 0.886405, "update_flag": false, "test_avg_loss": 7.489039, "test_total_loss": 89.868465, "test_acc": 0.7, "test_prec": 0.576923, "test_recall": 0.483871, "test_f1": 0.526316, "test_roc_auc": 0.67079, "test_pr_auc": 0.570273, "test_confusion_matrix": {"tn": 96, "fp": 22, "fn": 32, "tp": 30}, "test_mcc2": 0.311828, "test_mcc": 0.311828, "test_sn": 0.483871, "test_sp": 0.813559, "lr": 0.0001373083213083213, "cur_epoch_step": 2099, "train_global_avg_loss": 0.1880238002991812, "train_cur_epoch_loss": 97.13889496905279, "train_cur_epoch_avg_loss": 0.046278654106266214, "train_cur_epoch_time": 6352.7824993133545, "train_cur_epoch_avg_time": 3.026575750030183, "epoch": 16, "step": 33584} ################################################## Training, Epoch: 0017, Batch: 000416, Sample Num: 6656, Cur Loss: 0.06320013, Cur Avg Loss: 0.02643067, Log Avg loss: 0.03737966, Global Avg Loss: 0.18604666, Time: 3.0197 Steps: 34000, Updated lr: 0.000137 Training, Epoch: 0017, Batch: 001416, Sample Num: 22656, Cur Loss: 0.00003658, Cur Avg Loss: 0.03647448, Log Avg loss: 0.04065270, Global Avg Loss: 0.18189255, Time: 3.0127 Steps: 35000, Updated lr: 0.000135 ***** Running evaluation checkpoint-35683 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-35683 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6346.243845, Avg time per batch (s): 3.020000 {"eval_avg_loss": 1.168791, "eval_total_loss": 1046.067651, "eval_acc": 0.901565, "eval_prec": 0.879657, "eval_recall": 0.930418, "eval_f1": 0.904325, "eval_roc_auc": 0.957912, "eval_pr_auc": 0.952537, "eval_confusion_matrix": {"tn": 6246, "fp": 911, "fn": 498, "tp": 6659}, "eval_mcc2": 0.80447, "eval_mcc": 0.80447, "eval_sn": 0.930418, "eval_sp": 0.872712, "update_flag": false, "test_avg_loss": 7.905799, "test_total_loss": 94.869587, "test_acc": 0.716667, "test_prec": 0.596491, "test_recall": 0.548387, "test_f1": 0.571429, "test_roc_auc": 0.66088, "test_pr_auc": 0.564251, "test_confusion_matrix": {"tn": 95, "fp": 23, "fn": 28, "tp": 34}, "test_mcc2": 0.361079, "test_mcc": 0.361079, "test_sn": 0.548387, "test_sp": 0.805085, "lr": 0.00013326984126984127, "cur_epoch_step": 2099, "train_global_avg_loss": 0.17931444400904314, "train_cur_epoch_loss": 83.88599632698642, "train_cur_epoch_avg_loss": 0.03996474336683488, "train_cur_epoch_time": 6346.2438452243805, "train_cur_epoch_avg_time": 3.0234606218315294, "epoch": 17, "step": 35683} ################################################## Training, Epoch: 0018, Batch: 000317, Sample Num: 5072, Cur Loss: 0.00047175, Cur Avg Loss: 0.03229081, Log Avg loss: 0.04247433, Global Avg Loss: 0.17801982, Time: 3.0282 Steps: 36000, Updated lr: 0.000133 Training, Epoch: 0018, Batch: 001317, Sample Num: 21072, Cur Loss: 0.00711738, Cur Avg Loss: 0.03543305, Log Avg loss: 0.03642913, Global Avg Loss: 0.17419304, Time: 3.0164 Steps: 37000, Updated lr: 0.000131 ***** Running evaluation checkpoint-37782 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-37782 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6347.695591, Avg time per batch (s): 3.020000 {"eval_avg_loss": 1.49388, "eval_total_loss": 1337.022753, "eval_acc": 0.903381, "eval_prec": 0.894076, "eval_recall": 0.915188, "eval_f1": 0.904509, "eval_roc_auc": 0.956796, "eval_pr_auc": 0.950644, "eval_confusion_matrix": {"tn": 6381, "fp": 776, "fn": 607, "tp": 6550}, "eval_mcc2": 0.806988, "eval_mcc": 0.806988, "eval_sn": 0.915188, "eval_sp": 0.891575, "update_flag": false, "test_avg_loss": 9.914994, "test_total_loss": 118.979925, "test_acc": 0.705556, "test_prec": 0.591837, "test_recall": 0.467742, "test_f1": 0.522523, "test_roc_auc": 0.653157, "test_pr_auc": 0.534388, "test_confusion_matrix": {"tn": 98, "fp": 20, "fn": 33, "tp": 29}, "test_mcc2": 0.318408, "test_mcc": 0.318408, "test_sn": 0.467742, "test_sp": 0.830508, "lr": 0.00012923136123136123, "cur_epoch_step": 2099, "train_global_avg_loss": 0.17132437333312578, "train_cur_epoch_loss": 74.5001676974681, "train_cur_epoch_avg_loss": 0.03549317184252887, "train_cur_epoch_time": 6347.695591449738, "train_cur_epoch_avg_time": 3.024152258908879, "epoch": 18, "step": 37782} ################################################## Training, Epoch: 0019, Batch: 000218, Sample Num: 3488, Cur Loss: 0.00288612, Cur Avg Loss: 0.02458129, Log Avg loss: 0.03319357, Global Avg Loss: 0.17048253, Time: 3.0194 Steps: 38000, Updated lr: 0.000129 Training, Epoch: 0019, Batch: 001218, Sample Num: 19488, Cur Loss: 0.00000878, Cur Avg Loss: 0.03856746, Log Avg loss: 0.04161645, Global Avg Loss: 0.16717827, Time: 3.0042 Steps: 39000, Updated lr: 0.000127 ***** Running evaluation checkpoint-39881 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-39881 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6339.185900, Avg time per batch (s): 3.020000 {"eval_avg_loss": 1.284598, "eval_total_loss": 1149.715316, "eval_acc": 0.889688, "eval_prec": 0.855468, "eval_recall": 0.937823, "eval_f1": 0.894754, "eval_roc_auc": 0.950082, "eval_pr_auc": 0.941521, "eval_confusion_matrix": {"tn": 6023, "fp": 1134, "fn": 445, "tp": 6712}, "eval_mcc2": 0.783014, "eval_mcc": 0.783014, "eval_sn": 0.937823, "eval_sp": 0.841554, "update_flag": false, "test_avg_loss": 7.770908, "test_total_loss": 93.250891, "test_acc": 0.705556, "test_prec": 0.57377, "test_recall": 0.564516, "test_f1": 0.569106, "test_roc_auc": 0.678034, "test_pr_auc": 0.597428, "test_confusion_matrix": {"tn": 92, "fp": 26, "fn": 27, "tp": 35}, "test_mcc2": 0.345526, "test_mcc": 0.345526, "test_sn": 0.564516, "test_sp": 0.779661, "lr": 0.00012519288119288119, "cur_epoch_step": 2099, "train_global_avg_loss": 0.16432831102058687, "train_cur_epoch_loss": 80.59989853986539, "train_cur_epoch_avg_loss": 0.038399189394885846, "train_cur_epoch_time": 6339.185900211334, "train_cur_epoch_avg_time": 3.020098094431317, "epoch": 19, "step": 39881} ################################################## Training, Epoch: 0020, Batch: 000119, Sample Num: 1904, Cur Loss: 0.00000145, Cur Avg Loss: 0.02746861, Log Avg loss: 0.03689349, Global Avg Loss: 0.16392115, Time: 3.0274 Steps: 40000, Updated lr: 0.000125 Training, Epoch: 0020, Batch: 001119, Sample Num: 17904, Cur Loss: 0.00086853, Cur Avg Loss: 0.03250443, Log Avg loss: 0.03310369, Global Avg Loss: 0.16073048, Time: 3.0204 Steps: 41000, Updated lr: 0.000123 ***** Running evaluation checkpoint-41980 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-41980 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6345.156685, Avg time per batch (s): 3.020000 {"eval_avg_loss": 1.38216, "eval_total_loss": 1237.032799, "eval_acc": 0.903242, "eval_prec": 0.8848, "eval_recall": 0.927204, "eval_f1": 0.905506, "eval_roc_auc": 0.956405, "eval_pr_auc": 0.949898, "eval_confusion_matrix": {"tn": 6293, "fp": 864, "fn": 521, "tp": 6636}, "eval_mcc2": 0.807411, "eval_mcc": 0.807411, "eval_sn": 0.927204, "eval_sp": 0.879279, "update_flag": false, "test_avg_loss": 8.690414, "test_total_loss": 104.284968, "test_acc": 0.722222, "test_prec": 0.607143, "test_recall": 0.548387, "test_f1": 0.576271, "test_roc_auc": 0.68193, "test_pr_auc": 0.566108, "test_confusion_matrix": {"tn": 96, "fp": 22, "fn": 28, "tp": 34}, "test_mcc2": 0.371515, "test_mcc": 0.371515, "test_sn": 0.548387, "test_sp": 0.813559, "lr": 0.00012115440115440116, "cur_epoch_step": 2099, "train_global_avg_loss": 0.15771293232186762, "train_cur_epoch_loss": 67.21152705997005, "train_cur_epoch_avg_loss": 0.03202073704619821, "train_cur_epoch_time": 6345.156684875488, "train_cur_epoch_avg_time": 3.022942679788227, "epoch": 20, "step": 41980} ################################################## Training, Epoch: 0021, Batch: 000020, Sample Num: 320, Cur Loss: 0.00000930, Cur Avg Loss: 0.00838106, Log Avg loss: 0.03100669, Global Avg Loss: 0.15764182, Time: 3.0030 Steps: 42000, Updated lr: 0.000121 Training, Epoch: 0021, Batch: 001020, Sample Num: 16320, Cur Loss: 0.00007959, Cur Avg Loss: 0.01973351, Log Avg loss: 0.01996056, Global Avg Loss: 0.15443993, Time: 3.0004 Steps: 43000, Updated lr: 0.000119 Training, Epoch: 0021, Batch: 002020, Sample Num: 32320, Cur Loss: 0.00041480, Cur Avg Loss: 0.02546331, Log Avg loss: 0.03130770, Global Avg Loss: 0.15164147, Time: 3.0180 Steps: 44000, Updated lr: 0.000117 ***** Running evaluation checkpoint-44079 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-44079 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6335.278142, Avg time per batch (s): 3.020000 {"eval_avg_loss": 1.591753, "eval_total_loss": 1424.619202, "eval_acc": 0.912393, "eval_prec": 0.909078, "eval_recall": 0.916445, "eval_f1": 0.912747, "eval_roc_auc": 0.959022, "eval_pr_auc": 0.952979, "eval_confusion_matrix": {"tn": 6501, "fp": 656, "fn": 598, "tp": 6559}, "eval_mcc2": 0.824814, "eval_mcc": 0.824814, "eval_sn": 0.916445, "eval_sp": 0.908341, "update_flag": true, "test_avg_loss": 10.190666, "test_total_loss": 122.287997, "test_acc": 0.716667, "test_prec": 0.627907, "test_recall": 0.435484, "test_f1": 0.514286, "test_roc_auc": 0.673455, "test_pr_auc": 0.551604, "test_confusion_matrix": {"tn": 102, "fp": 16, "fn": 35, "tp": 27}, "test_mcc2": 0.334199, "test_mcc": 0.334199, "test_sn": 0.435484, "test_sp": 0.864407, "lr": 0.0001171159211159211, "cur_epoch_step": 2099, "train_global_avg_loss": 0.15141198223181662, "train_cur_epoch_loss": 53.29986592423655, "train_cur_epoch_avg_loss": 0.025392980430793972, "train_cur_epoch_time": 6335.278141975403, "train_cur_epoch_avg_time": 3.0182363706409734, "epoch": 21, "step": 44079} ################################################## Training, Epoch: 0022, Batch: 000921, Sample Num: 14736, Cur Loss: 0.00006795, Cur Avg Loss: 0.02382901, Log Avg loss: 0.02381051, Global Avg Loss: 0.14880078, Time: 3.0130 Steps: 45000, Updated lr: 0.000115 Training, Epoch: 0022, Batch: 001921, Sample Num: 30736, Cur Loss: 0.00000030, Cur Avg Loss: 0.02726233, Log Avg loss: 0.03042441, Global Avg Loss: 0.14622738, Time: 3.0329 Steps: 46000, Updated lr: 0.000113 ***** Running evaluation checkpoint-46178 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-46178 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6331.816589, Avg time per batch (s): 3.020000 {"eval_avg_loss": 1.674107, "eval_total_loss": 1498.326062, "eval_acc": 0.913162, "eval_prec": 0.914494, "eval_recall": 0.911555, "eval_f1": 0.913022, "eval_roc_auc": 0.959683, "eval_pr_auc": 0.952256, "eval_confusion_matrix": {"tn": 6547, "fp": 610, "fn": 633, "tp": 6524}, "eval_mcc2": 0.826328, "eval_mcc": 0.826328, "eval_sn": 0.911555, "eval_sp": 0.914769, "update_flag": true, "test_avg_loss": 11.10073, "test_total_loss": 133.208757, "test_acc": 0.727778, "test_prec": 0.644444, "test_recall": 0.467742, "test_f1": 0.542056, "test_roc_auc": 0.68357, "test_pr_auc": 0.575493, "test_confusion_matrix": {"tn": 102, "fp": 16, "fn": 33, "tp": 29}, "test_mcc2": 0.364499, "test_mcc": 0.364499, "test_sn": 0.467742, "test_sp": 0.864407, "lr": 0.00011307744107744109, "cur_epoch_step": 2099, "train_global_avg_loss": 0.14575301400786297, "train_cur_epoch_loss": 56.49391605883969, "train_cur_epoch_avg_loss": 0.026914681304830722, "train_cur_epoch_time": 6331.816588878632, "train_cur_epoch_avg_time": 3.0165872267168328, "epoch": 22, "step": 46178} ################################################## Training, Epoch: 0023, Batch: 000822, Sample Num: 13152, Cur Loss: 0.01068112, Cur Avg Loss: 0.01734952, Log Avg loss: 0.01838429, Global Avg Loss: 0.14350732, Time: 3.0039 Steps: 47000, Updated lr: 0.000111 Training, Epoch: 0023, Batch: 001822, Sample Num: 29152, Cur Loss: 0.00008815, Cur Avg Loss: 0.02015021, Log Avg loss: 0.02245238, Global Avg Loss: 0.14098534, Time: 3.0191 Steps: 48000, Updated lr: 0.000110 ***** Running evaluation checkpoint-48277 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-48277 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6329.194611, Avg time per batch (s): 3.020000 {"eval_avg_loss": 1.427807, "eval_total_loss": 1277.887415, "eval_acc": 0.891854, "eval_prec": 0.854686, "eval_recall": 0.94425, "eval_f1": 0.897238, "eval_roc_auc": 0.952182, "eval_pr_auc": 0.938896, "eval_confusion_matrix": {"tn": 6008, "fp": 1149, "fn": 399, "tp": 6758}, "eval_mcc2": 0.788047, "eval_mcc": 0.788047, "eval_sn": 0.94425, "eval_sp": 0.839458, "update_flag": false, "test_avg_loss": 8.686487, "test_total_loss": 104.237844, "test_acc": 0.705556, "test_prec": 0.569231, "test_recall": 0.596774, "test_f1": 0.582677, "test_roc_auc": 0.701886, "test_pr_auc": 0.593637, "test_confusion_matrix": {"tn": 90, "fp": 28, "fn": 25, "tp": 37}, "test_mcc2": 0.355642, "test_mcc": 0.355642, "test_sn": 0.596774, "test_sp": 0.762712, "lr": 0.00010903896103896105, "cur_epoch_step": 2099, "train_global_avg_loss": 0.14040758173676332, "train_cur_epoch_loss": 47.87414265061416, "train_cur_epoch_avg_loss": 0.02280807177256511, "train_cur_epoch_time": 6329.194610834122, "train_cur_epoch_avg_time": 3.0153380709071564, "epoch": 23, "step": 48277} ################################################## Training, Epoch: 0024, Batch: 000723, Sample Num: 11568, Cur Loss: 0.00000000, Cur Avg Loss: 0.02128667, Log Avg loss: 0.02655072, Global Avg Loss: 0.13864994, Time: 3.0009 Steps: 49000, Updated lr: 0.000108 Training, Epoch: 0024, Batch: 001723, Sample Num: 27568, Cur Loss: 0.03894496, Cur Avg Loss: 0.02144893, Log Avg loss: 0.02156625, Global Avg Loss: 0.13630827, Time: 3.0168 Steps: 50000, Updated lr: 0.000106 ***** Running evaluation checkpoint-50376 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-50376 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6333.098501, Avg time per batch (s): 3.020000 {"eval_avg_loss": 1.632077, "eval_total_loss": 1460.708468, "eval_acc": 0.905687, "eval_prec": 0.892843, "eval_recall": 0.922034, "eval_f1": 0.907204, "eval_roc_auc": 0.955785, "eval_pr_auc": 0.944254, "eval_confusion_matrix": {"tn": 6365, "fp": 792, "fn": 558, "tp": 6599}, "eval_mcc2": 0.811807, "eval_mcc": 0.811807, "eval_sn": 0.922034, "eval_sp": 0.889339, "update_flag": false, "test_avg_loss": 10.274511, "test_total_loss": 123.294137, "test_acc": 0.722222, "test_prec": 0.611111, "test_recall": 0.532258, "test_f1": 0.568966, "test_roc_auc": 0.688901, "test_pr_auc": 0.586529, "test_confusion_matrix": {"tn": 97, "fp": 21, "fn": 29, "tp": 33}, "test_mcc2": 0.36738, "test_mcc": 0.36738, "test_sn": 0.532258, "test_sp": 0.822034, "lr": 0.00010500048100048101, "cur_epoch_step": 2099, "train_global_avg_loss": 0.13546859355934285, "train_cur_epoch_loss": 45.90904563972731, "train_cur_epoch_avg_loss": 0.021871865478669514, "train_cur_epoch_time": 6333.098500967026, "train_cur_epoch_avg_time": 3.017197951866139, "epoch": 24, "step": 50376} ################################################## Training, Epoch: 0025, Batch: 000624, Sample Num: 9984, Cur Loss: 0.00000419, Cur Avg Loss: 0.01768885, Log Avg loss: 0.01999038, Global Avg Loss: 0.13402752, Time: 2.9958 Steps: 51000, Updated lr: 0.000104 Training, Epoch: 0025, Batch: 001624, Sample Num: 25984, Cur Loss: 0.12840496, Cur Avg Loss: 0.02215719, Log Avg loss: 0.02494544, Global Avg Loss: 0.13192979, Time: 3.0031 Steps: 52000, Updated lr: 0.000102 ***** Running evaluation checkpoint-52475 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-52475 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6334.109513, Avg time per batch (s): 3.020000 {"eval_avg_loss": 1.837776, "eval_total_loss": 1644.809378, "eval_acc": 0.913861, "eval_prec": 0.916596, "eval_recall": 0.910577, "eval_f1": 0.913577, "eval_roc_auc": 0.960507, "eval_pr_auc": 0.952798, "eval_confusion_matrix": {"tn": 6564, "fp": 593, "fn": 640, "tp": 6517}, "eval_mcc2": 0.827739, "eval_mcc": 0.827739, "eval_sn": 0.910577, "eval_sp": 0.917144, "update_flag": true, "test_avg_loss": 11.712074, "test_total_loss": 140.544891, "test_acc": 0.727778, "test_prec": 0.666667, "test_recall": 0.419355, "test_f1": 0.514851, "test_roc_auc": 0.696624, "test_pr_auc": 0.569111, "test_confusion_matrix": {"tn": 105, "fp": 13, "fn": 36, "tp": 26}, "test_mcc2": 0.356627, "test_mcc": 0.356627, "test_sn": 0.419355, "test_sp": 0.889831, "lr": 0.00010096200096200097, "cur_epoch_step": 2099, "train_global_avg_loss": 0.1310269454178799, "train_cur_epoch_loss": 51.273091657786544, "train_cur_epoch_avg_loss": 0.024427390022766338, "train_cur_epoch_time": 6334.10951256752, "train_cur_epoch_avg_time": 3.0176796153251644, "epoch": 25, "step": 52475} ################################################## Training, Epoch: 0026, Batch: 000525, Sample Num: 8400, Cur Loss: 0.00152775, Cur Avg Loss: 0.02215203, Log Avg loss: 0.02691963, Global Avg Loss: 0.12994847, Time: 3.0497 Steps: 53000, Updated lr: 0.000100 Training, Epoch: 0026, Batch: 001525, Sample Num: 24400, Cur Loss: 0.02238245, Cur Avg Loss: 0.02133665, Log Avg loss: 0.02090858, Global Avg Loss: 0.12792921, Time: 3.0155 Steps: 54000, Updated lr: 0.000098 ***** Running evaluation checkpoint-54574 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-54574 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6324.921898, Avg time per batch (s): 3.010000 {"eval_avg_loss": 1.628266, "eval_total_loss": 1457.297936, "eval_acc": 0.907573, "eval_prec": 0.892175, "eval_recall": 0.927204, "eval_f1": 0.909353, "eval_roc_auc": 0.958143, "eval_pr_auc": 0.947018, "eval_confusion_matrix": {"tn": 6355, "fp": 802, "fn": 521, "tp": 6636}, "eval_mcc2": 0.815775, "eval_mcc": 0.815775, "eval_sn": 0.927204, "eval_sp": 0.887942, "update_flag": false, "test_avg_loss": 10.885155, "test_total_loss": 130.621858, "test_acc": 0.711111, "test_prec": 0.592593, "test_recall": 0.516129, "test_f1": 0.551724, "test_roc_auc": 0.692865, "test_pr_auc": 0.581256, "test_confusion_matrix": {"tn": 96, "fp": 22, "fn": 30, "tp": 32}, "test_mcc2": 0.341868, "test_mcc": 0.341868, "test_sn": 0.516129, "test_sp": 0.813559, "lr": 9.692352092352093e-05, "cur_epoch_step": 2099, "train_global_avg_loss": 0.1267570626228595, "train_cur_epoch_loss": 42.00097477666871, "train_cur_epoch_avg_loss": 0.020009992747340978, "train_cur_epoch_time": 6324.921898126602, "train_cur_epoch_avg_time": 3.0133024764776573, "epoch": 26, "step": 54574} ################################################## Training, Epoch: 0027, Batch: 000426, Sample Num: 6816, Cur Loss: 0.00119002, Cur Avg Loss: 0.01227051, Log Avg loss: 0.01468982, Global Avg Loss: 0.12587031, Time: 3.0045 Steps: 55000, Updated lr: 0.000096 Training, Epoch: 0027, Batch: 001426, Sample Num: 22816, Cur Loss: 0.00003497, Cur Avg Loss: 0.01702687, Log Avg loss: 0.01905307, Global Avg Loss: 0.12396286, Time: 3.0198 Steps: 56000, Updated lr: 0.000094 ***** Running evaluation checkpoint-56673 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-56673 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6328.041125, Avg time per batch (s): 3.010000 {"eval_avg_loss": 1.487744, "eval_total_loss": 1331.530964, "eval_acc": 0.899329, "eval_prec": 0.870591, "eval_recall": 0.938103, "eval_f1": 0.903087, "eval_roc_auc": 0.951926, "eval_pr_auc": 0.93929, "eval_confusion_matrix": {"tn": 6159, "fp": 998, "fn": 443, "tp": 6714}, "eval_mcc2": 0.801071, "eval_mcc": 0.801071, "eval_sn": 0.938103, "eval_sp": 0.860556, "update_flag": false, "test_avg_loss": 9.666633, "test_total_loss": 115.999598, "test_acc": 0.694444, "test_prec": 0.557377, "test_recall": 0.548387, "test_f1": 0.552846, "test_roc_auc": 0.685962, "test_pr_auc": 0.567279, "test_confusion_matrix": {"tn": 91, "fp": 27, "fn": 28, "tp": 34}, "test_mcc2": 0.320826, "test_mcc": 0.320826, "test_sn": 0.548387, "test_sp": 0.771186, "lr": 9.288504088504088e-05, "cur_epoch_step": 2099, "train_global_avg_loss": 0.12271354209423664, "train_cur_epoch_loss": 36.90463552673053, "train_cur_epoch_avg_loss": 0.017582008350038366, "train_cur_epoch_time": 6328.041125059128, "train_cur_epoch_avg_time": 3.0147885302806707, "epoch": 27, "step": 56673} ################################################## Training, Epoch: 0028, Batch: 000327, Sample Num: 5232, Cur Loss: 0.00039962, Cur Avg Loss: 0.01436054, Log Avg loss: 0.01732022, Global Avg Loss: 0.12209194, Time: 3.0268 Steps: 57000, Updated lr: 0.000092 Training, Epoch: 0028, Batch: 001327, Sample Num: 21232, Cur Loss: 0.00000055, Cur Avg Loss: 0.01967705, Log Avg loss: 0.02141555, Global Avg Loss: 0.12035614, Time: 3.0126 Steps: 58000, Updated lr: 0.000090 ***** Running evaluation checkpoint-58772 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-58772 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6332.156321, Avg time per batch (s): 3.020000 {"eval_avg_loss": 2.245461, "eval_total_loss": 2009.687897, "eval_acc": 0.91407, "eval_prec": 0.927942, "eval_recall": 0.897862, "eval_f1": 0.912654, "eval_roc_auc": 0.961868, "eval_pr_auc": 0.956215, "eval_confusion_matrix": {"tn": 6658, "fp": 499, "fn": 731, "tp": 6426}, "eval_mcc2": 0.828576, "eval_mcc": 0.828576, "eval_sn": 0.897862, "eval_sp": 0.930278, "update_flag": false, "test_avg_loss": 13.163822, "test_total_loss": 157.965859, "test_acc": 0.727778, "test_prec": 0.658537, "test_recall": 0.435484, "test_f1": 0.524272, "test_roc_auc": 0.702501, "test_pr_auc": 0.599751, "test_confusion_matrix": {"tn": 104, "fp": 14, "fn": 35, "tp": 27}, "test_mcc2": 0.358986, "test_mcc": 0.358986, "test_sn": 0.435484, "test_sp": 0.881356, "lr": 8.884656084656085e-05, "cur_epoch_step": 2099, "train_global_avg_loss": 0.11900538812111428, "train_cur_epoch_loss": 39.64009954744699, "train_cur_epoch_avg_loss": 0.01888523084680657, "train_cur_epoch_time": 6332.156321287155, "train_cur_epoch_avg_time": 3.0167490811277538, "epoch": 28, "step": 58772} ################################################## Training, Epoch: 0029, Batch: 000228, Sample Num: 3648, Cur Loss: 0.00118103, Cur Avg Loss: 0.01246832, Log Avg loss: 0.01637143, Global Avg Loss: 0.11859369, Time: 3.0074 Steps: 59000, Updated lr: 0.000088 Training, Epoch: 0029, Batch: 001228, Sample Num: 19648, Cur Loss: 0.03583678, Cur Avg Loss: 0.01494835, Log Avg loss: 0.01551380, Global Avg Loss: 0.11687569, Time: 3.0128 Steps: 60000, Updated lr: 0.000086 ***** Running evaluation checkpoint-60871 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-60871 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6336.150163, Avg time per batch (s): 3.020000 {"eval_avg_loss": 1.998007, "eval_total_loss": 1788.215928, "eval_acc": 0.904499, "eval_prec": 0.895816, "eval_recall": 0.915467, "eval_f1": 0.905535, "eval_roc_auc": 0.955604, "eval_pr_auc": 0.946902, "eval_confusion_matrix": {"tn": 6395, "fp": 762, "fn": 605, "tp": 6552}, "eval_mcc2": 0.809193, "eval_mcc": 0.809193, "eval_sn": 0.915467, "eval_sp": 0.893531, "update_flag": false, "test_avg_loss": 12.701682, "test_total_loss": 152.420186, "test_acc": 0.666667, "test_prec": 0.520833, "test_recall": 0.403226, "test_f1": 0.454545, "test_roc_auc": 0.673729, "test_pr_auc": 0.585409, "test_confusion_matrix": {"tn": 95, "fp": 23, "fn": 37, "tp": 25}, "test_mcc2": 0.223841, "test_mcc": 0.223841, "test_sn": 0.403226, "test_sp": 0.805085, "lr": 8.480808080808081e-05, "cur_epoch_step": 2099, "train_global_avg_loss": 0.11550601798304184, "train_cur_epoch_loss": 36.78214999159759, "train_cur_epoch_avg_loss": 0.017523654117006952, "train_cur_epoch_time": 6336.150162696838, "train_cur_epoch_avg_time": 3.018651816434892, "epoch": 29, "step": 60871} ################################################## Training, Epoch: 0030, Batch: 000129, Sample Num: 2064, Cur Loss: 0.00000253, Cur Avg Loss: 0.00665389, Log Avg loss: 0.01928393, Global Avg Loss: 0.11527582, Time: 3.0212 Steps: 61000, Updated lr: 0.000085 Training, Epoch: 0030, Batch: 001129, Sample Num: 18064, Cur Loss: 0.00025862, Cur Avg Loss: 0.01354780, Log Avg loss: 0.01443711, Global Avg Loss: 0.11364939, Time: 3.0301 Steps: 62000, Updated lr: 0.000083 ***** Running evaluation checkpoint-62970 ***** Dev Dataset Instantaneous batch size per GPU = 16 Dev Dataset Num examples = 14314 ################################################## ***** Running testing checkpoint-62970 ***** Test Dataset Instantaneous batch size per GPU = 16 Test Dataset Num examples = [180] ################################################## Epoch Time: 6334.536252, Avg time per batch (s): 3.020000 {"eval_avg_loss": 2.214489, "eval_total_loss": 1981.9679, "eval_acc": 0.913232, "eval_prec": 0.922198, "eval_recall": 0.902613, "eval_f1": 0.912301, "eval_roc_auc": 0.960687, "eval_pr_auc": 0.955297, "eval_confusion_matrix": {"tn": 6612, "fp": 545, "fn": 697, "tp": 6460}, "eval_mcc2": 0.82665, "eval_mcc": 0.82665, "eval_sn": 0.902613, "eval_sp": 0.923851, "update_flag": false, "test_avg_loss": 13.623422, "test_total_loss": 163.481066, "test_acc": 0.733333, "test_prec": 0.659091, "test_recall": 0.467742, "test_f1": 0.54717, "test_roc_auc": 0.697991, "test_pr_auc": 0.601097, "test_confusion_matrix": {"tn": 103, "fp": 15, "fn": 33, "tp": 29}, "test_mcc2": 0.37663, "test_mcc": 0.37663, "test_sn": 0.467742, "test_sp": 0.872881, "lr": 8.076960076960077e-05, "cur_epoch_step": 2099, "train_global_avg_loss": 0.11210292435942668, "train_cur_epoch_loss": 28.154326267333165, "train_cur_epoch_avg_loss": 0.013413209274575115, "train_cur_epoch_time": 6334.536251544952, "train_cur_epoch_avg_time": 3.017882921174346, "epoch": 30, "step": 62970} ################################################## Training, Epoch: 0031, Batch: 000030, Sample Num: 480, Cur Loss: 0.03573383, Cur Avg Loss: 0.01341885, Log Avg loss: 0.01326143, Global Avg Loss: 0.11205593, Time: 3.0202 Steps: 63000, Updated lr: 0.000081 Training, Epoch: 0031, Batch: 001030, Sample Num: 16480, Cur Loss: 0.00000106, Cur Avg Loss: 0.01344852, Log Avg loss: 0.01344941, Global Avg Loss: 0.11051521, Time: 3.0477 Steps: 64000, Updated lr: 0.000079 Training, Epoch: 0031, Batch: 002030, Sample Num: 32480, Cur Loss: 0.00000528, Cur Avg Loss: 0.01382511, Log Avg loss: 0.01421299, Global Avg Loss: 0.10903363, Time: 3.0683 Steps: 65000, Updated lr: 0.000077