|
--- |
|
base_model: unsloth/mistral-7b-v0.3-bnb-4bit |
|
library_name: peft |
|
license: apache-2.0 |
|
tags: |
|
- unsloth |
|
- generated_from_trainer |
|
model-index: |
|
- name: Mistral-7B-v0.3_metamath_reverse |
|
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. --> |
|
|
|
# Mistral-7B-v0.3_metamath_reverse |
|
|
|
This model is a fine-tuned version of [unsloth/mistral-7b-v0.3-bnb-4bit](https://huggingface.co/unsloth/mistral-7b-v0.3-bnb-4bit) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 4.0369 |
|
|
|
## 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 | |
|
|:-------------:|:------:|:----:|:---------------:| |
|
| 0.7562 | 0.0211 | 13 | 8.8561 | |
|
| 8.5861 | 0.0421 | 26 | 6.7147 | |
|
| 6.683 | 0.0632 | 39 | 6.4347 | |
|
| 6.3623 | 0.0842 | 52 | 6.2959 | |
|
| 6.1966 | 0.1053 | 65 | 6.1023 | |
|
| 5.9253 | 0.1264 | 78 | 5.8562 | |
|
| 5.6996 | 0.1474 | 91 | 5.7402 | |
|
| 5.654 | 0.1685 | 104 | 5.5460 | |
|
| 5.4346 | 0.1896 | 117 | 5.3902 | |
|
| 5.2399 | 0.2106 | 130 | 5.1306 | |
|
| 5.1411 | 0.2317 | 143 | 5.0223 | |
|
| 5.0468 | 0.2527 | 156 | 4.9554 | |
|
| 4.9675 | 0.2738 | 169 | 4.8488 | |
|
| 4.8723 | 0.2949 | 182 | 4.9092 | |
|
| 4.9509 | 0.3159 | 195 | 4.6985 | |
|
| 4.7385 | 0.3370 | 208 | 4.7031 | |
|
| 4.631 | 0.3580 | 221 | 4.6471 | |
|
| 4.6294 | 0.3791 | 234 | 4.6124 | |
|
| 4.5562 | 0.4002 | 247 | 4.5880 | |
|
| 4.5684 | 0.4212 | 260 | 4.5116 | |
|
| 4.5965 | 0.4423 | 273 | 4.5065 | |
|
| 4.594 | 0.4633 | 286 | 4.4330 | |
|
| 4.5223 | 0.4844 | 299 | 4.4393 | |
|
| 4.4033 | 0.5055 | 312 | 4.4070 | |
|
| 4.3706 | 0.5265 | 325 | 4.3485 | |
|
| 4.3595 | 0.5476 | 338 | 4.3587 | |
|
| 4.3865 | 0.5687 | 351 | 4.2940 | |
|
| 4.342 | 0.5897 | 364 | 4.3082 | |
|
| 4.2976 | 0.6108 | 377 | 4.2683 | |
|
| 4.3627 | 0.6318 | 390 | 4.2331 | |
|
| 4.2364 | 0.6529 | 403 | 4.2331 | |
|
| 4.1543 | 0.6740 | 416 | 4.1827 | |
|
| 4.2475 | 0.6950 | 429 | 4.2243 | |
|
| 4.2247 | 0.7161 | 442 | 4.1690 | |
|
| 4.1115 | 0.7371 | 455 | 4.1257 | |
|
| 4.1388 | 0.7582 | 468 | 4.1157 | |
|
| 4.0912 | 0.7793 | 481 | 4.1659 | |
|
| 4.0903 | 0.8003 | 494 | 4.0926 | |
|
| 4.1036 | 0.8214 | 507 | 4.0859 | |
|
| 4.0692 | 0.8424 | 520 | 4.0732 | |
|
| 4.0634 | 0.8635 | 533 | 4.0823 | |
|
| 4.0463 | 0.8846 | 546 | 4.0597 | |
|
| 4.0948 | 0.9056 | 559 | 4.0447 | |
|
| 4.0496 | 0.9267 | 572 | 4.0293 | |
|
| 3.9855 | 0.9478 | 585 | 4.0449 | |
|
| 4.0289 | 0.9688 | 598 | 4.0360 | |
|
| 4.0147 | 0.9899 | 611 | 4.0369 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.12.0 |
|
- Transformers 4.44.0 |
|
- Pytorch 2.4.0+cu121 |
|
- Datasets 2.20.0 |
|
- Tokenizers 0.19.1 |
