| | --- |
| | license: mit |
| | tags: |
| | - generated_from_trainer |
| | datasets: |
| | - generator |
| | model-index: |
| | - name: gpt2-dp-2 |
| | 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-dp-2 |
| |
|
| | This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on the generator dataset. |
| | It achieves the following results on the evaluation set: |
| | - Loss: 4.3038 |
| |
|
| | ## 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: 0.0005 |
| | - train_batch_size: 64 |
| | - eval_batch_size: 64 |
| | - seed: 42 |
| | - gradient_accumulation_steps: 2 |
| | - 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: 1000 |
| | - num_epochs: 9 |
| | - mixed_precision_training: Native AMP |
| |
|
| | ### Training results |
| |
|
| | | Training Loss | Epoch | Step | Validation Loss | |
| | |:-------------:|:-----:|:----:|:---------------:| |
| | | 6.5574 | 0.53 | 500 | 5.4160 | |
| | | 5.0689 | 1.07 | 1000 | 4.9377 | |
| | | 4.6601 | 1.6 | 1500 | 4.6589 | |
| | | 4.3967 | 2.14 | 2000 | 4.4999 | |
| | | 4.1846 | 2.67 | 2500 | 4.3930 | |
| | | 4.0257 | 3.21 | 3000 | 4.3408 | |
| | | 3.8965 | 3.74 | 3500 | 4.2798 | |
| | | 3.7483 | 4.27 | 4000 | 4.2719 | |
| | | 3.6522 | 4.81 | 4500 | 4.2338 | |
| | | 3.4715 | 5.34 | 5000 | 4.2545 | |
| | | 3.4106 | 5.88 | 5500 | 4.2303 | |
| | | 3.2009 | 6.41 | 6000 | 4.2659 | |
| | | 3.1644 | 6.94 | 6500 | 4.2559 | |
| | | 2.9753 | 7.48 | 7000 | 4.2917 | |
| | | 2.9548 | 8.01 | 7500 | 4.2926 | |
| | | 2.846 | 8.55 | 8000 | 4.3038 | |
| |
|
| |
|
| | ### Framework versions |
| |
|
| | - Transformers 4.26.1 |
| | - Pytorch 1.11.0+cu113 |
| | - Datasets 2.13.0 |
| | - Tokenizers 0.13.3 |
| |
|
