<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
HBERTv1_emb_compress_48_L12_H512_A8
This model is a fine-tuned version of on the gokuls/wiki_book_corpus_complete_processed_bert_dataset dataset. It achieves the following results on the evaluation set:
- Loss: 6.0426
- Accuracy: 0.1512
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 56
- eval_batch_size: 56
- seed: 10
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 10000
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 7.102 | 0.1 | 10000 | 7.0825 | 0.0834 |
| 6.6803 | 0.19 | 20000 | 6.6756 | 0.1066 |
| 6.5348 | 0.29 | 30000 | 6.5298 | 0.1196 |
| 6.4394 | 0.38 | 40000 | 6.4389 | 0.1274 |
| 6.3686 | 0.48 | 50000 | 6.3726 | 0.1332 |
| 6.3213 | 0.57 | 60000 | 6.3189 | 0.1358 |
| 6.281 | 0.67 | 70000 | 6.2812 | 0.1382 |
| 6.2506 | 0.76 | 80000 | 6.2467 | 0.1401 |
| 6.221 | 0.86 | 90000 | 6.2216 | 0.1423 |
| 6.206 | 0.96 | 100000 | 6.1978 | 0.1431 |
| 6.1831 | 1.05 | 110000 | 6.1796 | 0.1449 |
| 6.1609 | 1.15 | 120000 | 6.1630 | 0.1457 |
| 6.153 | 1.24 | 130000 | 6.1505 | 0.1464 |
| 6.142 | 1.34 | 140000 | 6.1380 | 0.1471 |
| 6.1281 | 1.43 | 150000 | 6.1257 | 0.1477 |
| 6.1173 | 1.53 | 160000 | 6.1173 | 0.1481 |
| 6.1102 | 1.62 | 170000 | 6.1083 | 0.1489 |
| 6.1011 | 1.72 | 180000 | 6.1001 | 0.1487 |
| 6.0869 | 1.82 | 190000 | 6.0933 | 0.1493 |
| 6.0838 | 1.91 | 200000 | 6.0864 | 0.1494 |
| 6.0745 | 2.01 | 210000 | 6.0805 | 0.1499 |
| 6.0757 | 2.1 | 220000 | 6.0723 | 0.1503 |
| 6.0695 | 2.2 | 230000 | 6.0701 | 0.1502 |
| 6.0595 | 2.29 | 240000 | 6.0623 | 0.1506 |
| 6.0579 | 2.39 | 250000 | 6.0582 | 0.1506 |
| 6.0534 | 2.49 | 260000 | 6.0526 | 0.1509 |
| 6.0465 | 2.58 | 270000 | 6.0433 | 0.1510 |
Framework versions
- Transformers 4.33.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.14.5
- Tokenizers 0.13.3
