|
--- |
|
tags: |
|
- generated_from_trainer |
|
datasets: |
|
- gokuls/wiki_book_corpus_complete_processed_bert_dataset |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: HBERTv1_emb_compress_48_L10_H256_A4 |
|
results: |
|
- task: |
|
name: Masked Language Modeling |
|
type: fill-mask |
|
dataset: |
|
name: gokuls/wiki_book_corpus_complete_processed_bert_dataset |
|
type: gokuls/wiki_book_corpus_complete_processed_bert_dataset |
|
metrics: |
|
- name: Accuracy |
|
type: accuracy |
|
value: 0.15093352306316574 |
|
--- |
|
|
|
<!-- 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_L10_H256_A4 |
|
|
|
This model is a fine-tuned version of [](https://huggingface.co/) on the gokuls/wiki_book_corpus_complete_processed_bert_dataset dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 6.0495 |
|
- Accuracy: 0.1509 |
|
|
|
## 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: 64 |
|
- eval_batch_size: 64 |
|
- 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 | |
|
|:-------------:|:-----:|:------:|:---------------:|:--------:| |
|
| 7.1164 | 0.11 | 10000 | 7.0967 | 0.0830 | |
|
| 6.694 | 0.22 | 20000 | 6.6867 | 0.1065 | |
|
| 6.545 | 0.33 | 30000 | 6.5445 | 0.1171 | |
|
| 6.4556 | 0.44 | 40000 | 6.4527 | 0.1250 | |
|
| 6.3891 | 0.55 | 50000 | 6.3831 | 0.1305 | |
|
| 6.3404 | 0.66 | 60000 | 6.3334 | 0.1350 | |
|
| 6.2962 | 0.76 | 70000 | 6.2940 | 0.1377 | |
|
| 6.2669 | 0.87 | 80000 | 6.2629 | 0.1398 | |
|
| 6.2352 | 0.98 | 90000 | 6.2361 | 0.1412 | |
|
| 6.2179 | 1.09 | 100000 | 6.2150 | 0.1429 | |
|
| 6.191 | 1.2 | 110000 | 6.1970 | 0.1443 | |
|
| 6.1809 | 1.31 | 120000 | 6.1829 | 0.1441 | |
|
| 6.1699 | 1.42 | 130000 | 6.1692 | 0.1455 | |
|
| 6.1623 | 1.53 | 140000 | 6.1562 | 0.1453 | |
|
| 6.1422 | 1.64 | 150000 | 6.1480 | 0.1468 | |
|
| 6.1397 | 1.75 | 160000 | 6.1367 | 0.1468 | |
|
| 6.1342 | 1.86 | 170000 | 6.1284 | 0.1475 | |
|
| 6.1291 | 1.97 | 180000 | 6.1214 | 0.1478 | |
|
| 6.1157 | 2.08 | 190000 | 6.1132 | 0.1483 | |
|
| 6.1146 | 2.18 | 200000 | 6.1094 | 0.1484 | |
|
| 6.1018 | 2.29 | 210000 | 6.1013 | 0.1488 | |
|
| 6.1014 | 2.4 | 220000 | 6.0979 | 0.1488 | |
|
| 6.0935 | 2.51 | 230000 | 6.0936 | 0.1489 | |
|
| 6.0899 | 2.62 | 240000 | 6.0881 | 0.1491 | |
|
| 6.0858 | 2.73 | 250000 | 6.0851 | 0.1498 | |
|
| 6.0872 | 2.84 | 260000 | 6.0819 | 0.1497 | |
|
| 6.0858 | 2.95 | 270000 | 6.0784 | 0.1500 | |
|
| 6.0775 | 3.06 | 280000 | 6.0745 | 0.1501 | |
|
| 6.0715 | 3.17 | 290000 | 6.0720 | 0.1502 | |
|
| 6.0704 | 3.28 | 300000 | 6.0699 | 0.1502 | |
|
| 6.0678 | 3.39 | 310000 | 6.0668 | 0.1503 | |
|
| 6.0662 | 3.5 | 320000 | 6.0649 | 0.1503 | |
|
| 6.0569 | 3.6 | 330000 | 6.0622 | 0.1505 | |
|
| 6.0604 | 3.71 | 340000 | 6.0612 | 0.1506 | |
|
| 6.0525 | 3.82 | 350000 | 6.0586 | 0.1507 | |
|
| 6.0553 | 3.93 | 360000 | 6.0582 | 0.1506 | |
|
| 6.053 | 4.04 | 370000 | 6.0544 | 0.1508 | |
|
| 6.0594 | 4.15 | 380000 | 6.0553 | 0.1507 | |
|
| 6.0488 | 4.26 | 390000 | 6.0527 | 0.1509 | |
|
| 6.051 | 4.37 | 400000 | 6.0516 | 0.1509 | |
|
| 6.0553 | 4.48 | 410000 | 6.0518 | 0.1509 | |
|
| 6.0507 | 4.59 | 420000 | 6.0520 | 0.1509 | |
|
| 6.0514 | 4.7 | 430000 | 6.0501 | 0.1509 | |
|
| 6.0511 | 4.81 | 440000 | 6.0496 | 0.1511 | |
|
| 6.0527 | 4.92 | 450000 | 6.0493 | 0.1509 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.33.2 |
|
- Pytorch 1.14.0a0+410ce96 |
|
- Datasets 2.14.5 |
|
- Tokenizers 0.13.3 |
|
|
