|
|
--- |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
model-index: |
|
|
- name: GPT2-705M |
|
|
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. --> |
|
|
|
|
|
# GPT2-705M |
|
|
|
|
|
This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 5.4628 |
|
|
|
|
|
## 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: 0.00025 |
|
|
- train_batch_size: 16 |
|
|
- eval_batch_size: 8 |
|
|
- seed: 42 |
|
|
- gradient_accumulation_steps: 8 |
|
|
- total_train_batch_size: 128 |
|
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
|
- lr_scheduler_type: cosine |
|
|
- lr_scheduler_warmup_steps: 50 |
|
|
- num_epochs: 40 |
|
|
- mixed_precision_training: Native AMP |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|
|:-------------:|:-----:|:----:|:---------------:| |
|
|
| 9.7135 | 0.57 | 1 | 9.7272 | |
|
|
| 8.0222 | 1.71 | 3 | 9.3213 | |
|
|
| 7.6063 | 2.86 | 5 | 8.5841 | |
|
|
| 7.5596 | 4.0 | 7 | 7.9271 | |
|
|
| 7.4194 | 4.57 | 8 | 8.0942 | |
|
|
| 7.1644 | 5.71 | 10 | 7.5409 | |
|
|
| 6.8531 | 6.86 | 12 | 7.3028 | |
|
|
| 6.3614 | 8.0 | 14 | 9.3796 | |
|
|
| 8.5129 | 8.57 | 15 | 7.6361 | |
|
|
| 6.1325 | 9.71 | 17 | 6.7577 | |
|
|
| 5.8526 | 10.86 | 19 | 6.5249 | |
|
|
| 5.5941 | 12.0 | 21 | 6.2490 | |
|
|
| 5.4307 | 12.57 | 22 | 6.2442 | |
|
|
| 5.1381 | 13.71 | 24 | 5.9595 | |
|
|
| 4.8705 | 14.86 | 26 | 5.8944 | |
|
|
| 4.7083 | 16.0 | 28 | 5.7005 | |
|
|
| 4.5355 | 16.57 | 29 | 5.7459 | |
|
|
| 4.4187 | 17.71 | 31 | 5.5387 | |
|
|
| 4.3123 | 18.86 | 33 | 5.4863 | |
|
|
| 4.0269 | 20.0 | 35 | 5.3277 | |
|
|
| 3.942 | 20.57 | 36 | 5.3274 | |
|
|
| 3.784 | 21.71 | 38 | 5.3998 | |
|
|
| 3.4991 | 22.86 | 40 | 5.4628 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.39.1 |
|
|
- Pytorch 2.1.2+cu121 |
|
|
- Datasets 2.16.1 |
|
|
- Tokenizers 0.15.0 |
|
|
|
