|
--- |
|
license: apache-2.0 |
|
library_name: peft |
|
tags: |
|
- unsloth |
|
- generated_from_trainer |
|
base_model: mistralai/Mistral-7B-v0.3 |
|
model-index: |
|
- name: mistral_7b_v_MetaMathQA_40K_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_v_MetaMathQA_40K_reverse |
|
|
|
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.3](https://huggingface.co/mistralai/Mistral-7B-v0.3) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.4730 |
|
|
|
## 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: 0.0001 |
|
- 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: 0.02 |
|
- num_epochs: 1 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:------:|:----:|:---------------:| |
|
| 0.8354 | 0.0211 | 13 | 0.8810 | |
|
| 1.4577 | 0.0421 | 26 | 1.4281 | |
|
| 1.0366 | 0.0632 | 39 | 0.9662 | |
|
| 0.9024 | 0.0842 | 52 | 0.7634 | |
|
| 0.694 | 0.1053 | 65 | 0.7062 | |
|
| 0.665 | 0.1264 | 78 | 0.6924 | |
|
| 0.6381 | 0.1474 | 91 | 0.6665 | |
|
| 0.6481 | 0.1685 | 104 | 0.6725 | |
|
| 0.6394 | 0.1896 | 117 | 0.6697 | |
|
| 0.6486 | 0.2106 | 130 | 0.6728 | |
|
| 0.6381 | 0.2317 | 143 | 0.6631 | |
|
| 0.619 | 0.2527 | 156 | 0.6470 | |
|
| 0.6245 | 0.2738 | 169 | 0.6530 | |
|
| 0.6233 | 0.2949 | 182 | 0.6445 | |
|
| 0.6225 | 0.3159 | 195 | 0.6372 | |
|
| 0.6105 | 0.3370 | 208 | 0.6283 | |
|
| 0.5865 | 0.3580 | 221 | 0.6180 | |
|
| 0.5913 | 0.3791 | 234 | 0.6104 | |
|
| 0.5769 | 0.4002 | 247 | 0.6011 | |
|
| 0.586 | 0.4212 | 260 | 0.6021 | |
|
| 0.5945 | 0.4423 | 273 | 0.5921 | |
|
| 0.57 | 0.4633 | 286 | 0.5869 | |
|
| 0.5636 | 0.4844 | 299 | 0.5772 | |
|
| 0.5563 | 0.5055 | 312 | 0.5713 | |
|
| 0.5516 | 0.5265 | 325 | 0.5655 | |
|
| 0.5505 | 0.5476 | 338 | 0.5615 | |
|
| 0.5421 | 0.5687 | 351 | 0.5520 | |
|
| 0.5225 | 0.5897 | 364 | 0.5431 | |
|
| 0.5207 | 0.6108 | 377 | 0.5374 | |
|
| 0.5163 | 0.6318 | 390 | 0.5351 | |
|
| 0.5169 | 0.6529 | 403 | 0.5262 | |
|
| 0.5023 | 0.6740 | 416 | 0.5203 | |
|
| 0.483 | 0.6950 | 429 | 0.5153 | |
|
| 0.4999 | 0.7161 | 442 | 0.5074 | |
|
| 0.487 | 0.7371 | 455 | 0.5027 | |
|
| 0.4971 | 0.7582 | 468 | 0.4985 | |
|
| 0.4875 | 0.7793 | 481 | 0.4937 | |
|
| 0.4881 | 0.8003 | 494 | 0.4904 | |
|
| 0.4753 | 0.8214 | 507 | 0.4869 | |
|
| 0.4609 | 0.8424 | 520 | 0.4825 | |
|
| 0.4657 | 0.8635 | 533 | 0.4794 | |
|
| 0.4563 | 0.8846 | 546 | 0.4776 | |
|
| 0.4738 | 0.9056 | 559 | 0.4751 | |
|
| 0.4685 | 0.9267 | 572 | 0.4743 | |
|
| 0.4539 | 0.9478 | 585 | 0.4735 | |
|
| 0.4606 | 0.9688 | 598 | 0.4731 | |
|
| 0.457 | 0.9899 | 611 | 0.4730 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.7.1 |
|
- Transformers 4.40.2 |
|
- Pytorch 2.3.0+cu121 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.19.1 |
