| --- |
| 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: 5.5538 |
|
|
| ## 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: 50 |
| - mixed_precision_training: Native AMP |
|
|
| ### Training results |
|
|
| | Training Loss | Epoch | Step | Validation Loss | |
| |:-------------:|:-----:|:----:|:---------------:| |
| | 9.7407 | 0.57 | 1 | 9.7354 | |
| | 8.0949 | 1.71 | 3 | 9.2987 | |
| | 8.037 | 2.86 | 5 | 7.9942 | |
| | 8.4143 | 4.0 | 7 | 8.3825 | |
| | 7.7196 | 4.57 | 8 | 8.7978 | |
| | 7.2632 | 5.71 | 10 | 7.6261 | |
| | 6.9715 | 6.86 | 12 | 7.4135 | |
| | 6.4835 | 8.0 | 14 | 8.2776 | |
| | 7.1529 | 8.57 | 15 | 7.0085 | |
| | 6.1255 | 9.71 | 17 | 6.8228 | |
| | 5.9176 | 10.86 | 19 | 6.5603 | |
| | 5.5785 | 12.0 | 21 | 6.3862 | |
| | 5.4833 | 12.57 | 22 | 6.3011 | |
| | 5.1483 | 13.71 | 24 | 6.0480 | |
| | 4.9268 | 14.86 | 26 | 6.0532 | |
| | 4.6602 | 16.0 | 28 | 5.7750 | |
| | 4.5647 | 16.57 | 29 | 5.7046 | |
| | 4.3202 | 17.71 | 31 | 5.5333 | |
| | 4.1764 | 18.86 | 33 | 5.5809 | |
| | 4.1745 | 20.0 | 35 | 5.4089 | |
| | 4.0056 | 20.57 | 36 | 5.3978 | |
| | 3.8024 | 21.71 | 38 | 5.4085 | |
| | 3.5845 | 22.86 | 40 | 5.3279 | |
| | 3.4378 | 24.0 | 42 | 5.3881 | |
| | 3.3361 | 24.57 | 43 | 5.2754 | |
| | 3.2585 | 25.71 | 45 | 5.2913 | |
| | 3.168 | 26.86 | 47 | 5.4232 | |
| | 2.9045 | 28.0 | 49 | 5.5044 | |
| | 2.8709 | 28.57 | 50 | 5.5538 | |
|
|
|
|
| ### Framework versions |
|
|
| - Transformers 4.39.1 |
| - Pytorch 2.1.2+cu121 |
| - Datasets 2.16.1 |
| - Tokenizers 0.15.0 |
|
|
