|
--- |
|
license: apache-2.0 |
|
library_name: peft |
|
tags: |
|
- trl |
|
- sft |
|
- generated_from_trainer |
|
datasets: |
|
- generator |
|
base_model: mistralai/Mistral-7B-Instruct-v0.2 |
|
model-index: |
|
- name: mistral7binstruct_summarize |
|
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. --> |
|
|
|
# mistral7binstruct_summarize |
|
|
|
This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) on the generator dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.7060 |
|
|
|
## 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: 1 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: constant |
|
- training_steps: 500 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 1.7472 | 0.22 | 25 | 1.5467 | |
|
| 1.5628 | 0.43 | 50 | 1.4716 | |
|
| 1.5515 | 0.65 | 75 | 1.4513 | |
|
| 1.4966 | 0.87 | 100 | 1.4446 | |
|
| 1.4953 | 1.09 | 125 | 1.4509 | |
|
| 1.4828 | 1.3 | 150 | 1.4492 | |
|
| 1.4245 | 1.52 | 175 | 1.4513 | |
|
| 1.4047 | 1.74 | 200 | 1.4539 | |
|
| 1.4715 | 1.96 | 225 | 1.4541 | |
|
| 1.3183 | 2.17 | 250 | 1.4828 | |
|
| 1.2847 | 2.39 | 275 | 1.4780 | |
|
| 1.3934 | 2.61 | 300 | 1.4872 | |
|
| 1.3141 | 2.83 | 325 | 1.4947 | |
|
| 1.2404 | 3.04 | 350 | 1.5483 | |
|
| 1.1593 | 3.26 | 375 | 1.5560 | |
|
| 1.151 | 3.48 | 400 | 1.5593 | |
|
| 1.1381 | 3.7 | 425 | 1.5506 | |
|
| 1.2021 | 3.91 | 450 | 1.5614 | |
|
| 0.9731 | 4.13 | 475 | 1.6335 | |
|
| 0.9721 | 4.35 | 500 | 1.7060 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.9.0 |
|
- Transformers 4.38.2 |
|
- Pytorch 2.1.0+cu121 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.2 |
