File size: 2,299 Bytes
ad09416 88a1958 62e4c2c d9596f3 8e17eae 0fe6e72 78dbb4f d515e03 b91dbf4 6a2360d 56fb8aa 57ca7b4 6f14c22 929b2fd b61e7f6 f0efbae f51b257 9befc0e 5ffd261 6acb753 12b3bef ba5ce74 ba9c8c7 38a4779 0ecf70e c9b470c 62a72bb 7851d3f 98cbb31 8f6eed9 73998f1 c4069e2 2b98d0d fd08ee2 56de0cf cccdbb2 1a0205c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 |
Started at: 13:09:04
nb-bert-base, 5e-06, 256
({'_name_or_path': '/disk4/folder1/working/checkpoints/huggingface/native_pytorch/step4_8/', 'attention_probs_dropout_prob': 0.1, 'directionality': 'bidi', 'gradient_checkpointing': False, 'hidden_act': 'gelu', 'hidden_dropout_prob': 0.1, 'hidden_size': 768, 'initializer_range': 0.02, 'intermediate_size': 3072, 'layer_norm_eps': 1e-12, 'max_position_embeddings': 512, 'model_type': 'bert', 'num_attention_heads': 12, 'num_hidden_layers': 12, 'pad_token_id': 0, 'pooler_fc_size': 768, 'pooler_num_attention_heads': 12, 'pooler_num_fc_layers': 3, 'pooler_size_per_head': 128, 'pooler_type': 'first_token_transform', 'position_embedding_type': 'absolute', 'type_vocab_size': 2, 'vocab_size': 119547, '_commit_hash': '82b194c0b3ea1fcad65f1eceee04adb26f9f71ac'}, {})
Epoch: 0
Training loss: 0.5433284169435502 - MAE: 0.5775941266642696
Validation loss : 0.2084534830517239 - MAE: 0.34436377863461687
Epoch: 1
Training loss: 0.18881268858909606 - MAE: 0.3319951213456936
Validation loss : 0.17533069021171993 - MAE: 0.31952199275705395
Epoch: 2
Training loss: 0.17236681401729584 - MAE: 0.31305812195355187
Validation loss : 0.16540057957172394 - MAE: 0.30917810361613507
Epoch: 3
Training loss: 0.16512651681900026 - MAE: 0.30562505926503336
Validation loss : 0.1583948234717051 - MAE: 0.2991497249751852
Epoch: 4
Training loss: 0.15748567909002303 - MAE: 0.2989958723461345
Validation loss : 0.15284320215384165 - MAE: 0.29026856200804096
Epoch: 5
Training loss: 0.14964943289756774 - MAE: 0.2896343934318654
Validation loss : 0.14675226559241614 - MAE: 0.2830523245210482
Epoch: 6
Training loss: 0.1409749785065651 - MAE: 0.2818433281871407
Validation loss : 0.14069517536295784 - MAE: 0.2762524538160992
Epoch: 7
Training loss: 0.1349916523694992 - MAE: 0.27349521608195476
Validation loss : 0.13396325210730234 - MAE: 0.26954649511104584
Epoch: 8
Training loss: 0.12729438930749892 - MAE: 0.2662118251014631
Validation loss : 0.12773916290866005 - MAE: 0.2634949845222456
Epoch: 9
Training loss: 0.1206126120686531 - MAE: 0.25962342232666624
Validation loss : 0.12221866349379222 - MAE: 0.2589877239039535
Epoch: 10
Training loss: 0.11287372469902039 - MAE: 0.2514929032844295
Validation loss : 0.12077544712358051 - MAE: 0.25974176291313755
Epoch: 11
|