metadata
license: mit
base_model: gpt2
tags:
- generated_from_trainer
model-index:
- name: gpt2-p10k-cossine
results: []
gpt2-p10k-cossine
This model is a fine-tuned version of gpt2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0234
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: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| No log | 0.2 | 200 | 0.0386 |
| No log | 0.4 | 400 | 0.0523 |
| 0.1425 | 0.6 | 600 | 1.0542 |
| 0.1425 | 0.8 | 800 | 1.0459 |
| 0.9177 | 1.0 | 1000 | 0.3677 |
| 0.9177 | 1.2 | 1200 | 0.0296 |
| 0.9177 | 1.4 | 1400 | 0.0284 |
| 0.0421 | 1.6 | 1600 | 0.0275 |
| 0.0421 | 1.8 | 1800 | 0.0274 |
| 0.029 | 2.0 | 2000 | 0.0264 |
| 0.029 | 2.2 | 2200 | 0.0258 |
| 0.029 | 2.4 | 2400 | 0.0256 |
| 0.0276 | 2.6 | 2600 | 0.0254 |
| 0.0276 | 2.8 | 2800 | 0.0252 |
| 0.0265 | 3.0 | 3000 | 0.0251 |
| 0.0265 | 3.2 | 3200 | 0.0247 |
| 0.0265 | 3.4 | 3400 | 0.0247 |
| 0.0256 | 3.6 | 3600 | 0.0246 |
| 0.0256 | 3.8 | 3800 | 0.0252 |
| 0.0262 | 4.0 | 4000 | 0.0249 |
| 0.0262 | 4.2 | 4200 | 0.0244 |
| 0.0262 | 4.4 | 4400 | 0.0242 |
| 0.0255 | 4.6 | 4600 | 0.0242 |
| 0.0255 | 4.8 | 4800 | 0.0248 |
| 0.0251 | 5.0 | 5000 | 0.0240 |
| 0.0251 | 5.2 | 5200 | 0.0242 |
| 0.0251 | 5.4 | 5400 | 0.0239 |
| 0.0254 | 5.6 | 5600 | 0.0239 |
| 0.0254 | 5.8 | 5800 | 0.0237 |
| 0.0244 | 6.0 | 6000 | 0.0237 |
| 0.0244 | 6.2 | 6200 | 0.0239 |
| 0.0244 | 6.4 | 6400 | 0.0237 |
| 0.0244 | 6.6 | 6600 | 0.0237 |
| 0.0244 | 6.8 | 6800 | 0.0238 |
| 0.0246 | 7.0 | 7000 | 0.0236 |
| 0.0246 | 7.2 | 7200 | 0.0235 |
| 0.0246 | 7.4 | 7400 | 0.0235 |
| 0.0242 | 7.6 | 7600 | 0.0235 |
| 0.0242 | 7.8 | 7800 | 0.0235 |
| 0.0244 | 8.0 | 8000 | 0.0236 |
| 0.0244 | 8.2 | 8200 | 0.0234 |
| 0.0244 | 8.4 | 8400 | 0.0235 |
| 0.0246 | 8.6 | 8600 | 0.0234 |
| 0.0246 | 8.8 | 8800 | 0.0234 |
| 0.0237 | 9.0 | 9000 | 0.0234 |
| 0.0237 | 9.2 | 9200 | 0.0234 |
| 0.0237 | 9.4 | 9400 | 0.0234 |
| 0.0241 | 9.6 | 9600 | 0.0234 |
| 0.0241 | 9.8 | 9800 | 0.0234 |
| 0.024 | 10.0 | 10000 | 0.0234 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.3.0
- Datasets 2.19.1
- Tokenizers 0.19.1