|
--- |
|
base_model: unsloth/llama-3-8b |
|
library_name: peft |
|
license: llama3 |
|
tags: |
|
- unsloth |
|
- generated_from_trainer |
|
model-index: |
|
- name: Meta-Llama-3-8B_metamath_default |
|
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. --> |
|
|
|
# Meta-Llama-3-8B_metamath_default |
|
|
|
This model is a fine-tuned version of [unsloth/llama-3-8b](https://huggingface.co/unsloth/llama-3-8b) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.5068 |
|
|
|
## 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: 0.0003 |
|
- 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_ratio: 0.02 |
|
- num_epochs: 1 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:------:|:----:|:---------------:| |
|
| 0.8666 | 0.0211 | 13 | 0.7509 | |
|
| 0.6952 | 0.0421 | 26 | 0.7368 | |
|
| 0.7079 | 0.0632 | 39 | 0.7196 | |
|
| 0.6922 | 0.0842 | 52 | 0.7066 | |
|
| 0.6565 | 0.1053 | 65 | 0.7074 | |
|
| 0.6791 | 0.1264 | 78 | 0.7263 | |
|
| 0.6858 | 0.1474 | 91 | 0.7019 | |
|
| 0.6693 | 0.1685 | 104 | 0.6926 | |
|
| 0.6503 | 0.1896 | 117 | 0.6922 | |
|
| 0.6488 | 0.2106 | 130 | 0.6925 | |
|
| 0.6505 | 0.2317 | 143 | 0.6844 | |
|
| 0.6533 | 0.2527 | 156 | 0.6842 | |
|
| 0.6505 | 0.2738 | 169 | 0.6709 | |
|
| 0.6456 | 0.2949 | 182 | 0.6661 | |
|
| 0.6307 | 0.3159 | 195 | 0.6699 | |
|
| 0.6144 | 0.3370 | 208 | 0.6629 | |
|
| 0.6286 | 0.3580 | 221 | 0.6547 | |
|
| 0.6261 | 0.3791 | 234 | 0.6469 | |
|
| 0.6365 | 0.4002 | 247 | 0.6482 | |
|
| 0.6108 | 0.4212 | 260 | 0.6428 | |
|
| 0.6207 | 0.4423 | 273 | 0.6322 | |
|
| 0.6219 | 0.4633 | 286 | 0.6265 | |
|
| 0.6133 | 0.4844 | 299 | 0.6213 | |
|
| 0.5944 | 0.5055 | 312 | 0.6138 | |
|
| 0.5871 | 0.5265 | 325 | 0.6034 | |
|
| 0.5827 | 0.5476 | 338 | 0.6013 | |
|
| 0.5714 | 0.5687 | 351 | 0.5923 | |
|
| 0.5512 | 0.5897 | 364 | 0.5849 | |
|
| 0.5636 | 0.6108 | 377 | 0.5755 | |
|
| 0.5564 | 0.6318 | 390 | 0.5684 | |
|
| 0.5444 | 0.6529 | 403 | 0.5647 | |
|
| 0.5431 | 0.6740 | 416 | 0.5582 | |
|
| 0.5311 | 0.6950 | 429 | 0.5533 | |
|
| 0.5323 | 0.7161 | 442 | 0.5458 | |
|
| 0.5172 | 0.7371 | 455 | 0.5386 | |
|
| 0.5113 | 0.7582 | 468 | 0.5341 | |
|
| 0.4989 | 0.7793 | 481 | 0.5296 | |
|
| 0.4929 | 0.8003 | 494 | 0.5264 | |
|
| 0.5266 | 0.8214 | 507 | 0.5214 | |
|
| 0.5075 | 0.8424 | 520 | 0.5184 | |
|
| 0.4917 | 0.8635 | 533 | 0.5150 | |
|
| 0.5078 | 0.8846 | 546 | 0.5124 | |
|
| 0.4897 | 0.9056 | 559 | 0.5099 | |
|
| 0.4879 | 0.9267 | 572 | 0.5081 | |
|
| 0.5007 | 0.9478 | 585 | 0.5073 | |
|
| 0.4979 | 0.9688 | 598 | 0.5071 | |
|
| 0.4991 | 0.9899 | 611 | 0.5068 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.12.0 |
|
- Transformers 4.44.0 |
|
- Pytorch 2.4.0+cu121 |
|
- Datasets 2.20.0 |
|
- Tokenizers 0.19.1 |
