File size: 2,077 Bytes
24b3561 44ffac2 bc13da5 dca4917 2b40235 c7e9b30 6e6c50d 1b9a17f 00fe6aa 5df5650 56b2efd 60bb8f2 76c4dcc c3761e5 5cae6d8 6578a93 545dd44 b01437c edb36a0 20a7976 e663e95 679c5b6 541a874 9c25485 3b62d95 9a1155e 7ddea04 86b170e b7501c4 3146ae8 b64b8e5 8ba3907 |
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 |
Started at: 10:02:30
nb-bert-base, 0.001, 128
({'_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: 0.33461407989263536 - MAE: 0.4395451926176116
Validation loss : 0.17470807801274693 - MAE: 0.316590567311014
Epoch: 1
Training loss: 0.167546104490757 - MAE: 0.3104947695666053
Validation loss : 0.1605402261895292 - MAE: 0.3041267300080164
Epoch: 2
Training loss: 0.15953998982906342 - MAE: 0.30230224392344485
Validation loss : 0.156581671360661 - MAE: 0.3003702843252824
Epoch: 3
Training loss: 0.15586254999041557 - MAE: 0.2976227208408245
Validation loss : 0.15435482824549956 - MAE: 0.29782302866508426
Epoch: 4
Training loss: 0.153594990670681 - MAE: 0.2951023367494838
Validation loss : 0.15327392255558686 - MAE: 0.29758807291884654
Epoch: 5
Training loss: 0.1521633943915367 - MAE: 0.293013468516263
Validation loss : 0.1528188367100323 - MAE: 0.29786525073805276
Epoch: 6
Training loss: 0.15114636063575745 - MAE: 0.29207026473332626
Validation loss : 0.1511023048968876 - MAE: 0.2940484620567765
Epoch: 7
Training loss: 0.14836823791265488 - MAE: 0.289356569209421
Validation loss : 0.15079707608503454 - MAE: 0.2948241208277115
Epoch: 8
Training loss: 0.14847442269325256 - MAE: 0.2894169052100116
Validation loss : 0.15011538915774403 - MAE: 0.2942575668480824
Epoch: 9
Training loss: 0.1463117568194866 - MAE: 0.2868930664850179
|