|
--- |
|
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: 3.5063 |
|
|
|
## 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: 15 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 6.8119 | 1.0 | 3 | 6.8091 | |
|
| 6.6598 | 2.0 | 6 | 6.8246 | |
|
| 6.0219 | 3.0 | 9 | 6.2434 | |
|
| 5.1608 | 4.0 | 12 | 5.4866 | |
|
| 4.6874 | 5.0 | 15 | 5.7119 | |
|
| 4.7554 | 6.0 | 18 | 4.9916 | |
|
| 4.3244 | 7.0 | 21 | 4.8076 | |
|
| 4.3358 | 8.0 | 24 | 4.7170 | |
|
| 4.3353 | 9.0 | 27 | 4.4035 | |
|
| 4.0477 | 10.0 | 30 | 4.1959 | |
|
| 3.7513 | 11.0 | 33 | 3.9729 | |
|
| 3.7101 | 12.0 | 36 | 3.8325 | |
|
| 3.333 | 13.0 | 39 | 3.7540 | |
|
| 3.3225 | 14.0 | 42 | 3.6116 | |
|
| 2.9902 | 15.0 | 45 | 3.5063 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.39.1 |
|
- Pytorch 2.1.2+cu121 |
|
- Datasets 2.16.1 |
|
- Tokenizers 0.15.0 |
|
|
