run2-babylm-original-2345-gpt2-small
This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.5191
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: 0.0002
- train_batch_size: 64
- eval_batch_size: 64
- seed: 2345
- gradient_accumulation_steps: 8
- total_train_batch_size: 512
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine_with_min_lr
- lr_scheduler_warmup_steps: 100
- num_epochs: 20.0
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 4.2704 | 0.5091 | 160 | 4.1863 |
| 3.8840 | 1.0159 | 320 | 3.8323 |
| 3.6971 | 1.5251 | 480 | 3.6240 |
| 3.4703 | 2.0318 | 640 | 3.4051 |
| 3.2992 | 2.5410 | 800 | 3.2471 |
| 3.1849 | 3.0477 | 960 | 3.1311 |
| 3.0794 | 3.5569 | 1120 | 3.0315 |
| 2.9876 | 4.0636 | 1280 | 2.9488 |
| 2.9447 | 4.5728 | 1440 | 2.8787 |
| 2.8554 | 5.0796 | 1600 | 2.8259 |
| 2.8436 | 5.5887 | 1760 | 2.7830 |
| 2.7865 | 6.0955 | 1920 | 2.7498 |
| 2.7874 | 6.6046 | 2080 | 2.7208 |
| 2.7438 | 7.1114 | 2240 | 2.6972 |
| 2.7024 | 7.6205 | 2400 | 2.6767 |
| 2.6741 | 8.1273 | 2560 | 2.6586 |
| 2.6616 | 8.6364 | 2720 | 2.6429 |
| 2.6123 | 9.1432 | 2880 | 2.6285 |
| 2.6142 | 9.6523 | 3040 | 2.6162 |
| 2.6184 | 10.1591 | 3200 | 2.6051 |
| 2.5942 | 10.6683 | 3360 | 2.5947 |
| 2.5524 | 11.1750 | 3520 | 2.5863 |
| 2.5718 | 11.6842 | 3680 | 2.5774 |
| 2.5861 | 12.1909 | 3840 | 2.5708 |
| 2.5396 | 12.7001 | 4000 | 2.5629 |
| 2.5229 | 13.2068 | 4160 | 2.5584 |
| 2.5371 | 13.7160 | 4320 | 2.5519 |
| 2.5274 | 14.2228 | 4480 | 2.5478 |
| 2.5441 | 14.7319 | 4640 | 2.5435 |
| 2.5171 | 15.2387 | 4800 | 2.5396 |
| 2.5446 | 15.7478 | 4960 | 2.5356 |
| 2.5116 | 16.2546 | 5120 | 2.5337 |
| 2.5138 | 16.7637 | 5280 | 2.5301 |
| 2.4947 | 17.2705 | 5440 | 2.5280 |
| 2.4837 | 17.7796 | 5600 | 2.5253 |
| 2.4999 | 18.2864 | 5760 | 2.5242 |
| 2.5032 | 18.7955 | 5920 | 2.5221 |
| 2.4722 | 19.3023 | 6080 | 2.5211 |
| 2.4781 | 19.8115 | 6240 | 2.5195 |
| 2.5099 | 20.0 | 6300 | 2.5191 |
Framework versions
- Transformers 5.13.0
- Pytorch 2.11.0+cu128
- Datasets 5.0.0
- Tokenizers 0.22.2