|
--- |
|
license: apache-2.0 |
|
base_model: mistralai/Mistral-7B-v0.1 |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: out |
|
results: [] |
|
--- |
|
|
|
[<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl) |
|
# mistral-alpaca-finetune |
|
|
|
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the mhenrichsen/alpaca_2k_test dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.9808 |
|
|
|
## 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-06 |
|
- train_batch_size: 16 |
|
- eval_batch_size: 16 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: cosine |
|
- lr_scheduler_warmup_steps: 10 |
|
- num_epochs: 3 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 0.9152 | 0.01 | 1 | 0.9037 | |
|
| 0.9101 | 0.15 | 18 | 0.8461 | |
|
| 0.7589 | 0.3 | 36 | 0.8437 | |
|
| 0.8274 | 0.45 | 54 | 0.8441 | |
|
| 0.7255 | 0.61 | 72 | 0.8435 | |
|
| 0.85 | 0.76 | 90 | 0.8419 | |
|
| 0.9083 | 0.91 | 108 | 0.8408 | |
|
| 0.3208 | 1.06 | 126 | 0.9177 | |
|
| 0.3738 | 1.21 | 144 | 0.8924 | |
|
| 0.4034 | 1.36 | 162 | 0.8914 | |
|
| 0.3936 | 1.51 | 180 | 0.9032 | |
|
| 0.3188 | 1.66 | 198 | 0.9001 | |
|
| 0.4331 | 1.82 | 216 | 0.8973 | |
|
| 0.3946 | 1.97 | 234 | 0.8963 | |
|
| 0.1531 | 2.12 | 252 | 0.9653 | |
|
| 0.1741 | 2.27 | 270 | 0.9841 | |
|
| 0.2371 | 2.42 | 288 | 0.9784 | |
|
| 0.271 | 2.57 | 306 | 0.9801 | |
|
| 0.2632 | 2.72 | 324 | 0.9808 | |
|
| 0.1691 | 2.87 | 342 | 0.9808 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.35.2 |
|
- Pytorch 2.0.1+cu118 |
|
- Datasets 2.15.0 |
|
- Tokenizers 0.15.0 |
|
|
