File size: 2,163 Bytes
f39f508 fbfbf9b 81f45fb a5681e3 2aaff5f 82f9ed0 cf6c108 63f986f d0e5ef7 de4b210 210f1a3 2e9e16b 1de0954 91e41f8 6faa25e 1e92c55 dd8041b f054aa5 dfa0ed0 8ca1d95 6b5474c 0003f3f 6e10179 db71dfb 61fc0e0 2bf8603 235efc3 5c23fdf 810b005 ec25e3b 29dc690 810b21c 3b3df70 516f1d7 |
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 |
Started at: 16:47:13
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: 0.3211674815416336 - MAE: 0.43050889442825496
Validation loss : 0.17011426476871266 - MAE: 0.3182868112479769
Epoch: 1
Training loss: 0.1676739512383938 - MAE: 0.3128174295269125
Validation loss : 0.15711577148998485 - MAE: 0.30288625474990954
Epoch: 2
Training loss: 0.16074925452470779 - MAE: 0.30505208991793226
Validation loss : 0.15224149253438501 - MAE: 0.29741199591796025
Epoch: 3
Training loss: 0.15376607656478883 - MAE: 0.2960283221636555
Validation loss : 0.14967771794866114 - MAE: 0.29525018144397497
Epoch: 4
Training loss: 0.15107476785779 - MAE: 0.29479635191266035
Validation loss : 0.14622102195725722 - MAE: 0.28988063518034757
Epoch: 5
Training loss: 0.14845454767346383 - MAE: 0.2908094051374329
Validation loss : 0.14483271539211273 - MAE: 0.28878485930326336
Epoch: 6
Training loss: 0.1469292964041233 - MAE: 0.2894193336372531
Validation loss : 0.14388507078675664 - MAE: 0.28755274267308994
Epoch: 7
Training loss: 0.14546455055475235 - MAE: 0.2893174664939615
Validation loss : 0.14310433540274115 - MAE: 0.2867540477886474
Epoch: 8
Training loss: 0.14304511204361917 - MAE: 0.28596896352655493
Validation loss : 0.14294052080196493 - MAE: 0.2872696199373502
Epoch: 9
Training loss: 0.1416772884130478 - MAE: 0.2841209235577174
Validation loss : 0.14274786281235077 - MAE: 0.2866903104994522
Epoch: 10
|