|
--- |
|
license: apache-2.0 |
|
base_model: mistralai/Mistral-7B-Instruct-v0.1 |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: radia-fine-tune-mistral-7b-lora-v2 |
|
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. --> |
|
|
|
# radia-fine-tune-mistral-7b-lora-v2 |
|
|
|
This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.5165 |
|
|
|
## 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: 8 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 2 |
|
- total_train_batch_size: 16 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: constant |
|
- lr_scheduler_warmup_ratio: 0.3 |
|
- num_epochs: 1 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 1.0388 | 0.09 | 5 | 0.7940 | |
|
| 0.8312 | 0.17 | 10 | 0.7249 | |
|
| 0.6977 | 0.26 | 15 | 0.6799 | |
|
| 0.6852 | 0.34 | 20 | 0.6458 | |
|
| 0.6458 | 0.43 | 25 | 0.6152 | |
|
| 0.6263 | 0.52 | 30 | 0.5930 | |
|
| 0.5636 | 0.6 | 35 | 0.5717 | |
|
| 0.573 | 0.69 | 40 | 0.5538 | |
|
| 0.5467 | 0.78 | 45 | 0.5382 | |
|
| 0.5212 | 0.86 | 50 | 0.5277 | |
|
| 0.4598 | 0.95 | 55 | 0.5165 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.36.0.dev0 |
|
- Pytorch 2.1.0+cu118 |
|
- Datasets 2.15.0 |
|
- Tokenizers 0.15.0 |
|
|
