Finetune_FULL_16_8_RTE.log

01/10/2021 11:03:19 1
01/10/2021 11:03:19 Launching the MT-DNN training
01/10/2021 11:03:19 Loading data/canonical_data/bert_base_uncased_lower/rte_train.json as task 1
01/10/2021 11:03:19 ####################
01/10/2021 11:03:19 {'log_file': 'checkpoints/finetune-rte-full16-$/log.log', 'tensorboard': False, 'tensorboard_logdir': 'tensorboard_logdir', 'init_checkpoint': 'checkpoints/mtdnn-full16/model_4.pt', 'data_dir': 'data/canonical_data/bert_base_uncased_lower', 'data_sort_on': False, 'name': 'farmer', 'task_def': 'experiments/glue/glue_task_def.yml', 'train_datasets': ['rte'], 'test_datasets': ['rte'], 'glue_format_on': False, 'mkd_opt': 0, 'do_padding': False, 'update_bert_opt': 0, 'multi_gpu_on': True, 'mem_cum_type': 'simple', 'answer_num_turn': 5, 'answer_mem_drop_p': 0.1, 'answer_att_hidden_size': 128, 'answer_att_type': 'bilinear', 'answer_rnn_type': 'gru', 'answer_sum_att_type': 'bilinear', 'answer_merge_opt': 1, 'answer_mem_type': 1, 'max_answer_len': 10, 'answer_dropout_p': 0.1, 'answer_weight_norm_on': False, 'dump_state_on': False, 'answer_opt': 1, 'mtl_opt': 0, 'ratio': 0, 'mix_opt': 0, 'max_seq_len': 512, 'init_ratio': 1, 'encoder_type': <EncoderModelType.BERT: 1>, 'num_hidden_layers': -1, 'bert_model_type': 'bert-base-uncased', 'do_lower_case': False, 'masked_lm_prob': 0.15, 'short_seq_prob': 0.2, 'max_predictions_per_seq': 128, 'bin_on': False, 'bin_size': 64, 'bin_grow_ratio': 0.5, 'cuda': True, 'log_per_updates': 500, 'save_per_updates': 10000, 'save_per_updates_on': False, 'epochs': 30, 'batch_size': 8, 'batch_size_eval': 8, 'optimizer': 'adamax', 'grad_clipping': 0.0, 'global_grad_clipping': 1.0, 'weight_decay': 0, 'learning_rate': 5e-05, 'momentum': 0, 'warmup': 0.1, 'warmup_schedule': 'warmup_linear', 'adam_eps': 1e-06, 'vb_dropout': True, 'dropout_p': 0.1, 'dropout_w': 0.0, 'bert_dropout_p': 0.1, 'model_ckpt': 'checkpoints/model_0.pt', 'resume': False, 'have_lr_scheduler': True, 'multi_step_lr': '10,20,30', 'lr_gamma': 0.5, 'scheduler_type': 'ms', 'output_dir': 'checkpoints/finetune-rte-full16-$', 'seed': 2018, 'grad_accumulation_step': 1, 'fp16': False, 'fp16_opt_level': 'O1', 'adv_train': False, 'adv_opt': 0, 'adv_norm_level': 0, 'adv_p_norm': 'inf', 'adv_alpha': 1, 'adv_k': 1, 'adv_step_size': 0.001, 'adv_noise_var': 1e-05, 'adv_epsilon': 1e-06, 'loss_pred': True, 'collect_uncertainty': None, 'collect_topk': 0.1, 'load_ranked_data': None, 'mc_dropout': 0, 'finetune': True, 'encode_mode': False, 'task_def_list': [{'self': '{}', 'label_vocab': '<data_utils.vocab.Vocabulary object at 0x7fbea53758b0>', 'n_class': '2', 'data_type': '<DataFormat.PremiseAndOneHypothesis: 2>', 'task_type': '<TaskType.Classification: 1>', 'metric_meta': '(<Metric.ACC: 0>,)', 'split_names': "['train', 'dev', 'test']", 'enable_san': 'False', 'dropout_p': 'None', 'loss': '<LossCriterion.CeCriterion: 0>', 'kd_loss': '<LossCriterion.MseCriterion: 1>', 'adv_loss': '<LossCriterion.SymKlCriterion: 7>', '__class__': "<class 'experiments.exp_def.TaskDef'>"}]}
01/10/2021 11:03:19 ####################
01/10/2021 11:03:19 ############# Gradient Accumulation Info #############
01/10/2021 11:03:19 number of step: 9360
01/10/2021 11:03:19 number of grad grad_accumulation step: 1
01/10/2021 11:03:19 adjusted number of step: 9360
01/10/2021 11:03:19 ############# Gradient Accumulation Info #############
01/10/2021 11:03:29 
############# Model Arch of MT-DNN #############
SANBertNetwork(
  (dropout_list): ModuleList(
    (0): DropoutWrapper()
    (1): DropoutWrapper()
    (2): DropoutWrapper()
    (3): DropoutWrapper()
    (4): DropoutWrapper()
    (5): DropoutWrapper()
    (6): DropoutWrapper()
    (7): DropoutWrapper()
  )
  (bert): BertModel(
    (embeddings): BertEmbeddings(
      (word_embeddings): Embedding(30522, 768, padding_idx=0)
      (position_embeddings): Embedding(512, 768)
      (token_type_embeddings): Embedding(2, 768)
      (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
      (dropout): Dropout(p=0.1, inplace=False)
    )
    (encoder): BertEncoder(
      (layer): ModuleList(
        (0): BertLayer(
          (attention): BertAttention(
            (self): BertSelfAttention(
              (query): Linear(in_features=768, out_features=768, bias=True)
              (key): Linear(in_features=768, out_features=768, bias=True)
              (value): Linear(in_features=768, out_features=768, bias=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
            (output): BertSelfOutput(
              (dense): Linear(in_features=768, out_features=768, bias=True)
              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
          )
          (intermediate): BertIntermediate(
            (dense): Linear(in_features=768, out_features=3072, bias=True)
          )
          (output): BertOutput(
            (dense): Linear(in_features=3072, out_features=768, bias=True)
            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
            (dropout): Dropout(p=0.1, inplace=False)
          )
        )
        (1): BertLayer(
          (attention): BertAttention(
            (self): BertSelfAttention(
              (query): Linear(in_features=768, out_features=768, bias=True)
              (key): Linear(in_features=768, out_features=768, bias=True)
              (value): Linear(in_features=768, out_features=768, bias=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
            (output): BertSelfOutput(
              (dense): Linear(in_features=768, out_features=768, bias=True)
              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
          )
          (intermediate): BertIntermediate(
            (dense): Linear(in_features=768, out_features=3072, bias=True)
          )
          (output): BertOutput(
            (dense): Linear(in_features=3072, out_features=768, bias=True)
            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
            (dropout): Dropout(p=0.1, inplace=False)
          )
        )
        (2): BertLayer(
          (attention): BertAttention(
            (self): BertSelfAttention(
              (query): Linear(in_features=768, out_features=768, bias=True)
              (key): Linear(in_features=768, out_features=768, bias=True)
              (value): Linear(in_features=768, out_features=768, bias=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
            (output): BertSelfOutput(
              (dense): Linear(in_features=768, out_features=768, bias=True)
              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
          )
          (intermediate): BertIntermediate(
            (dense): Linear(in_features=768, out_features=3072, bias=True)
          )
          (output): BertOutput(
            (dense): Linear(in_features=3072, out_features=768, bias=True)
            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
            (dropout): Dropout(p=0.1, inplace=False)
          )
        )
        (3): BertLayer(
          (attention): BertAttention(
            (self): BertSelfAttention(
              (query): Linear(in_features=768, out_features=768, bias=True)
              (key): Linear(in_features=768, out_features=768, bias=True)
              (value): Linear(in_features=768, out_features=768, bias=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
            (output): BertSelfOutput(
              (dense): Linear(in_features=768, out_features=768, bias=True)
              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
          )
          (intermediate): BertIntermediate(
            (dense): Linear(in_features=768, out_features=3072, bias=True)
          )
          (output): BertOutput(
            (dense): Linear(in_features=3072, out_features=768, bias=True)
            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
            (dropout): Dropout(p=0.1, inplace=False)
          )
        )
        (4): BertLayer(
          (attention): BertAttention(
            (self): BertSelfAttention(
              (query): Linear(in_features=768, out_features=768, bias=True)
              (key): Linear(in_features=768, out_features=768, bias=True)
              (value): Linear(in_features=768, out_features=768, bias=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
            (output): BertSelfOutput(
              (dense): Linear(in_features=768, out_features=768, bias=True)
              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
          )
          (intermediate): BertIntermediate(
            (dense): Linear(in_features=768, out_features=3072, bias=True)
          )
          (output): BertOutput(
            (dense): Linear(in_features=3072, out_features=768, bias=True)
            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
            (dropout): Dropout(p=0.1, inplace=False)
          )
        )
        (5): BertLayer(
          (attention): BertAttention(
            (self): BertSelfAttention(
              (query): Linear(in_features=768, out_features=768, bias=True)
              (key): Linear(in_features=768, out_features=768, bias=True)
              (value): Linear(in_features=768, out_features=768, bias=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
            (output): BertSelfOutput(
              (dense): Linear(in_features=768, out_features=768, bias=True)
              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
          )
          (intermediate): BertIntermediate(
            (dense): Linear(in_features=768, out_features=3072, bias=True)
          )
          (output): BertOutput(
            (dense): Linear(in_features=3072, out_features=768, bias=True)
            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
            (dropout): Dropout(p=0.1, inplace=False)
          )
        )
        (6): BertLayer(
          (attention): BertAttention(
            (self): BertSelfAttention(
              (query): Linear(in_features=768, out_features=768, bias=True)
              (key): Linear(in_features=768, out_features=768, bias=True)
              (value): Linear(in_features=768, out_features=768, bias=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
            (output): BertSelfOutput(
              (dense): Linear(in_features=768, out_features=768, bias=True)
              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
          )
          (intermediate): BertIntermediate(
            (dense): Linear(in_features=768, out_features=3072, bias=True)
          )
          (output): BertOutput(
            (dense): Linear(in_features=3072, out_features=768, bias=True)
            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
            (dropout): Dropout(p=0.1, inplace=False)
          )
        )
        (7): BertLayer(
          (attention): BertAttention(
            (self): BertSelfAttention(
              (query): Linear(in_features=768, out_features=768, bias=True)
              (key): Linear(in_features=768, out_features=768, bias=True)
              (value): Linear(in_features=768, out_features=768, bias=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
            (output): BertSelfOutput(
              (dense): Linear(in_features=768, out_features=768, bias=True)
              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
          )
          (intermediate): BertIntermediate(
            (dense): Linear(in_features=768, out_features=3072, bias=True)
          )
          (output): BertOutput(
            (dense): Linear(in_features=3072, out_features=768, bias=True)
            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
            (dropout): Dropout(p=0.1, inplace=False)
          )
        )
        (8): BertLayer(
          (attention): BertAttention(
            (self): BertSelfAttention(
              (query): Linear(in_features=768, out_features=768, bias=True)
              (key): Linear(in_features=768, out_features=768, bias=True)
              (value): Linear(in_features=768, out_features=768, bias=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
            (output): BertSelfOutput(
              (dense): Linear(in_features=768, out_features=768, bias=True)
              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
          )
          (intermediate): BertIntermediate(
            (dense): Linear(in_features=768, out_features=3072, bias=True)
          )
          (output): BertOutput(
            (dense): Linear(in_features=3072, out_features=768, bias=True)
            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
            (dropout): Dropout(p=0.1, inplace=False)
          )
        )
        (9): BertLayer(
          (attention): BertAttention(
            (self): BertSelfAttention(
              (query): Linear(in_features=768, out_features=768, bias=True)
              (key): Linear(in_features=768, out_features=768, bias=True)
              (value): Linear(in_features=768, out_features=768, bias=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
            (output): BertSelfOutput(
              (dense): Linear(in_features=768, out_features=768, bias=True)
              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
          )
          (intermediate): BertIntermediate(
            (dense): Linear(in_features=768, out_features=3072, bias=True)
          )
          (output): BertOutput(
            (dense): Linear(in_features=3072, out_features=768, bias=True)
            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
            (dropout): Dropout(p=0.1, inplace=False)
          )
        )
        (10): BertLayer(
          (attention): BertAttention(
            (self): BertSelfAttention(
              (query): Linear(in_features=768, out_features=768, bias=True)
              (key): Linear(in_features=768, out_features=768, bias=True)
              (value): Linear(in_features=768, out_features=768, bias=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
            (output): BertSelfOutput(
              (dense): Linear(in_features=768, out_features=768, bias=True)
              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
          )
          (intermediate): BertIntermediate(
            (dense): Linear(in_features=768, out_features=3072, bias=True)
          )
          (output): BertOutput(
            (dense): Linear(in_features=3072, out_features=768, bias=True)
            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
            (dropout): Dropout(p=0.1, inplace=False)
          )
        )
        (11): BertLayer(
          (attention): BertAttention(
            (self): BertSelfAttention(
              (query): Linear(in_features=768, out_features=768, bias=True)
              (key): Linear(in_features=768, out_features=768, bias=True)
              (value): Linear(in_features=768, out_features=768, bias=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
            (output): BertSelfOutput(
              (dense): Linear(in_features=768, out_features=768, bias=True)
              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
          )
          (intermediate): BertIntermediate(
            (dense): Linear(in_features=768, out_features=3072, bias=True)
          )
          (output): BertOutput(
            (dense): Linear(in_features=3072, out_features=768, bias=True)
            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
            (dropout): Dropout(p=0.1, inplace=False)
          )
        )
      )
    )
    (pooler): BertPooler(
      (dense): Linear(in_features=768, out_features=768, bias=True)
      (activation): Tanh()
    )
  )
  (loss_pred_fc): Linear(in_features=768, out_features=1, bias=True)
  (scoring_list): ModuleList(
    (0): Linear(in_features=768, out_features=3, bias=True)
    (1): Linear(in_features=768, out_features=2, bias=True)
    (2): Linear(in_features=768, out_features=2, bias=True)
    (3): Linear(in_features=768, out_features=2, bias=True)
    (4): Linear(in_features=768, out_features=2, bias=True)
    (5): Linear(in_features=768, out_features=2, bias=True)
    (6): Linear(in_features=768, out_features=2, bias=True)
    (7): Linear(in_features=768, out_features=1, bias=True)
  )
)

01/10/2021 11:03:29 Total number of params: 109495313
01/10/2021 11:03:29 At epoch 0
01/10/2021 11:03:29 Task [ 1] updates[     0] train loss[0.03986] remaining[0:01:07]
01/10/2021 11:04:09 Task rte -- epoch 0 -- Dev ACC: 79.783
01/10/2021 11:04:17 [new test scores saved.]
01/10/2021 11:04:22 At epoch 1
01/10/2021 11:05:03 Task rte -- epoch 1 -- Dev ACC: 79.061
01/10/2021 11:05:11 [new test scores saved.]
01/10/2021 11:05:16 At epoch 2
01/10/2021 11:05:57 Task rte -- epoch 2 -- Dev ACC: 79.783
01/10/2021 11:06:06 [new test scores saved.]
01/10/2021 11:06:12 At epoch 3
01/10/2021 11:06:21 Task [ 1] updates[   500] train loss[0.18329] remaining[0:00:32]
01/10/2021 11:06:53 Task rte -- epoch 3 -- Dev ACC: 77.978
01/10/2021 11:07:03 [new test scores saved.]
01/10/2021 11:07:08 At epoch 4
01/10/2021 11:07:50 Task rte -- epoch 4 -- Dev ACC: 80.866
01/10/2021 11:08:00 [new test scores saved.]
01/10/2021 11:08:09 At epoch 5
01/10/2021 11:08:51 Task rte -- epoch 5 -- Dev ACC: 82.671
01/10/2021 11:09:01 [new test scores saved.]
01/10/2021 11:09:07 At epoch 6
01/10/2021 11:09:23 Task [ 1] updates[  1000] train loss[0.12695] remaining[0:00:23]
01/10/2021 11:09:49 Task rte -- epoch 6 -- Dev ACC: 80.144
01/10/2021 11:09:58 [new test scores saved.]
01/10/2021 11:10:04 At epoch 7
01/10/2021 11:10:46 Task rte -- epoch 7 -- Dev ACC: 80.144
01/10/2021 11:10:56 [new test scores saved.]
01/10/2021 11:11:04 At epoch 8
01/10/2021 11:11:45 Task rte -- epoch 8 -- Dev ACC: 79.783
01/10/2021 11:11:54 [new test scores saved.]
01/10/2021 11:11:59 At epoch 9
01/10/2021 11:12:23 Task [ 1] updates[  1500] train loss[0.09790] remaining[0:00:14]
01/10/2021 11:12:39 Task rte -- epoch 9 -- Dev ACC: 79.783
01/10/2021 11:12:49 [new test scores saved.]
01/10/2021 11:12:54 At epoch 10
01/10/2021 11:13:36 Task rte -- epoch 10 -- Dev ACC: 79.783
01/10/2021 11:13:45 [new test scores saved.]
01/10/2021 11:13:52 At epoch 11
01/10/2021 11:14:34 Task rte -- epoch 11 -- Dev ACC: 80.866
01/10/2021 11:14:44 [new test scores saved.]
01/10/2021 11:14:49 At epoch 12
01/10/2021 11:15:23 Task [ 1] updates[  2000] train loss[0.07657] remaining[0:00:07]
01/10/2021 11:15:32 Task rte -- epoch 12 -- Dev ACC: 80.505
01/10/2021 11:15:41 [new test scores saved.]
01/10/2021 11:15:48 At epoch 13
01/10/2021 11:16:31 Task rte -- epoch 13 -- Dev ACC: 79.783
01/10/2021 11:16:40 [new test scores saved.]
01/10/2021 11:16:48 At epoch 14
01/10/2021 11:17:30 Task rte -- epoch 14 -- Dev ACC: 80.866
01/10/2021 11:17:40 [new test scores saved.]
01/10/2021 11:17:47 At epoch 15
01/10/2021 11:18:30 Task rte -- epoch 15 -- Dev ACC: 80.866
01/10/2021 11:18:40 [new test scores saved.]
01/10/2021 11:18:46 At epoch 16
01/10/2021 11:18:47 Task [ 1] updates[  2500] train loss[0.06459] remaining[0:00:44]
01/10/2021 11:19:28 Task rte -- epoch 16 -- Dev ACC: 82.310
01/10/2021 11:19:38 [new test scores saved.]
01/10/2021 11:19:48 At epoch 17
01/10/2021 11:20:30 Task rte -- epoch 17 -- Dev ACC: 80.505
01/10/2021 11:20:40 [new test scores saved.]
01/10/2021 11:20:47 At epoch 18
01/10/2021 11:21:32 Task rte -- epoch 18 -- Dev ACC: 81.588
01/10/2021 11:21:43 [new test scores saved.]
01/10/2021 11:21:52 At epoch 19
01/10/2021 11:22:02 Task [ 1] updates[  3000] train loss[0.05637] remaining[0:00:34]
01/10/2021 11:22:36 Task rte -- epoch 19 -- Dev ACC: 80.866
01/10/2021 11:22:45 [new test scores saved.]
01/10/2021 11:22:53 At epoch 20
01/10/2021 11:23:36 Task rte -- epoch 20 -- Dev ACC: 81.588
01/10/2021 11:23:45 [new test scores saved.]
01/10/2021 11:23:55 At epoch 21
01/10/2021 11:24:37 Task rte -- epoch 21 -- Dev ACC: 82.310
01/10/2021 11:24:47 [new test scores saved.]
01/10/2021 11:24:55 At epoch 22
01/10/2021 11:25:12 Task [ 1] updates[  3500] train loss[0.04955] remaining[0:00:22]
01/10/2021 11:25:36 Task rte -- epoch 22 -- Dev ACC: 80.505
01/10/2021 11:25:45 [new test scores saved.]
01/10/2021 11:25:51 At epoch 23
01/10/2021 11:26:32 Task rte -- epoch 23 -- Dev ACC: 81.227
01/10/2021 11:26:42 [new test scores saved.]
01/10/2021 11:26:48 At epoch 24
01/10/2021 11:27:31 Task rte -- epoch 24 -- Dev ACC: 81.588
01/10/2021 11:27:40 [new test scores saved.]
01/10/2021 11:27:46 At epoch 25
01/10/2021 11:28:12 Task [ 1] updates[  4000] train loss[0.04400] remaining[0:00:14]
01/10/2021 11:28:29 Task rte -- epoch 25 -- Dev ACC: 80.144
01/10/2021 11:28:38 [new test scores saved.]
01/10/2021 11:28:44 At epoch 26
01/10/2021 11:29:28 Task rte -- epoch 26 -- Dev ACC: 80.866
01/10/2021 11:29:38 [new test scores saved.]
01/10/2021 11:29:44 At epoch 27
01/10/2021 11:30:27 Task rte -- epoch 27 -- Dev ACC: 80.505
01/10/2021 11:30:36 [new test scores saved.]
01/10/2021 11:30:43 At epoch 28
01/10/2021 11:31:18 Task [ 1] updates[  4500] train loss[0.03947] remaining[0:00:06]
01/10/2021 11:31:25 Task rte -- epoch 28 -- Dev ACC: 80.505
01/10/2021 11:31:34 [new test scores saved.]
01/10/2021 11:31:40 At epoch 29
01/10/2021 11:32:21 Task rte -- epoch 29 -- Dev ACC: 80.144
01/10/2021 11:32:31 [new test scores saved.]