|
--- |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- bleu |
|
model-index: |
|
- name: BERT2BERT_finetuned |
|
results: [] |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# BERT2BERT_finetuned |
|
|
|
This model is a fine-tuned version of [JulienRPA/BERT2BERT_pretrained_LC-QuAD_2.0](https://huggingface.co/JulienRPA/BERT2BERT_pretrained_LC-QuAD_2.0) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.3523 |
|
- Bleu: 95.2821 |
|
- Em: 0.1415 |
|
- Rm: 0.3046 |
|
- Gen Len: 58.7746 |
|
|
|
## 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: 5e-05 |
|
- train_batch_size: 32 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 2000 |
|
- num_epochs: 300.0 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Bleu | Em | Gen Len | Validation Loss | Rm | |
|
|:-------------:|:------:|:-----:|:-------:|:------:|:-------:|:---------------:|:------:| |
|
| 4.0279 | 12.82 | 500 | 49.841 | 0.0 | 51.6403 | 2.4031 | 0.0 | |
|
| 1.3442 | 25.64 | 1000 | 85.0177 | 0.0 | 57.9784 | 0.5014 | 0.0 | |
|
| 0.2522 | 38.46 | 1500 | 94.0714 | 0.0168 | 57.9137 | 0.3293 | 0.0216 | |
|
| 0.1534 | 51.28 | 2000 | 94.4328 | 0.0024 | 58.9448 | 0.3207 | 0.0072 | |
|
| 0.1305 | 64.1 | 2500 | 94.0708 | 0.0 | 59.6115 | 0.3247 | 0.0 | |
|
| 0.1226 | 76.92 | 3000 | 94.3143 | 0.0024 | 58.235 | 0.3325 | 0.0024 | |
|
| 0.1131 | 89.74 | 3500 | 94.5678 | 0.0048 | 59.6811 | 0.3401 | 0.0144 | |
|
| 0.1053 | 102.56 | 4000 | 94.4738 | 0.0168 | 59.0288 | 0.3374 | 0.0552 | |
|
| 0.0999 | 115.38 | 4500 | 94.6291 | 0.0336 | 58.6283 | 0.3437 | 0.0624 | |
|
| 0.0941 | 128.21 | 5000 | 94.7896 | 0.0695 | 58.4149 | 0.3512 | 0.1271 | |
|
| 0.0904 | 141.03 | 5500 | 94.4101 | 0.0719 | 58.2518 | 0.3424 | 0.1439 | |
|
| 0.0833 | 153.85 | 6000 | 94.7141 | 0.0887 | 59.0312 | 0.3462 | 0.1775 | |
|
| 0.0772 | 166.67 | 6500 | 94.6758 | 0.0911 | 59.0767 | 0.3467 | 0.2062 | |
|
| 0.0722 | 179.49 | 7000 | 94.5698 | 0.1055 | 58.1415 | 0.3462 | 0.2398 | |
|
| 0.0669 | 192.31 | 7500 | 95.0365 | 0.1223 | 58.7794 | 0.3537 | 0.2782 | |
|
| 0.062 | 205.13 | 8000 | 94.8694 | 0.1247 | 58.211 | 0.3505 | 0.2686 | |
|
| 0.0576 | 217.95 | 8500 | 94.8168 | 0.1271 | 59.0791 | 0.3511 | 0.2926 | |
|
| 0.0539 | 230.77 | 9000 | 95.1935 | 0.1367 | 58.6787 | 0.3490 | 0.3046 | |
|
| 0.0502 | 243.59 | 9500 | 95.1882 | 0.1319 | 58.5228 | 0.3490 | 0.3141 | |
|
| 0.0473 | 256.41 | 10000 | 95.1198 | 0.1319 | 58.4245 | 0.3504 | 0.307 | |
|
| 0.045 | 269.23 | 10500 | 0.3505 | 95.047 | 0.1343 | 0.307 | 58.3213| |
|
| 0.0429 | 282.05 | 11000 | 0.3522 | 95.2397| 0.1391 | 0.3046 | 58.7242| |
|
| 0.0416 | 294.87 | 11500 | 0.3523 | 95.2821| 0.1415 | 0.3046 | 58.7746| |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.30.0.dev0 |
|
- Pytorch 2.0.1+cu118 |
|
- Datasets 2.12.0 |
|
- Tokenizers 0.13.3 |
|
|
