metadata
license: mit
base_model: gpt2-large
tags:
- generated_from_trainer
model-index:
- name: tiq
results: []
tiq
This model is a fine-tuned version of gpt2-large on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 5.5477
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 200
- num_epochs: 1
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 6.2342 | 0.04 | 100 | 6.1857 |
| 5.7599 | 0.07 | 200 | 5.7751 |
| 5.7433 | 0.11 | 300 | 5.7142 |
| 5.6021 | 0.15 | 400 | 5.6776 |
| 5.5084 | 0.18 | 500 | 5.6349 |
| 5.3825 | 0.22 | 600 | 5.6201 |
| 5.6698 | 0.26 | 700 | 5.5831 |
| 5.4089 | 0.29 | 800 | 5.5687 |
| 5.601 | 0.33 | 900 | 5.5574 |
| 5.4708 | 0.37 | 1000 | 5.5555 |
| 5.5956 | 0.4 | 1100 | 5.5520 |
| 5.4704 | 0.44 | 1200 | 5.5494 |
| 5.4824 | 0.47 | 1300 | 5.5502 |
| 5.589 | 0.51 | 1400 | 5.5478 |
| 5.5612 | 0.55 | 1500 | 5.5456 |
| 5.4741 | 0.58 | 1600 | 5.5430 |
| 5.463 | 0.62 | 1700 | 5.5426 |
| 5.5071 | 0.66 | 1800 | 5.5424 |
| 5.5469 | 0.69 | 1900 | 5.5419 |
| 5.4266 | 0.73 | 2000 | 5.5428 |
| 5.4848 | 0.77 | 2100 | 5.5438 |
| 5.5069 | 0.8 | 2200 | 5.5446 |
| 5.5885 | 0.84 | 2300 | 5.5469 |
| 5.4484 | 0.88 | 2400 | 5.5462 |
| 5.3859 | 0.91 | 2500 | 5.5475 |
| 5.465 | 0.95 | 2600 | 5.5476 |
| 5.4355 | 0.99 | 2700 | 5.5477 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.2.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2