metadata
license: apache-2.0
base_model: mistralai/Mistral-7B-Instruct-v0.2
tags:
- trl
- sft
- generated_from_trainer
model-index:
- name: results
results: []
results
This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0307
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: 4
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- lr_scheduler_warmup_steps: 0.03
- training_steps: 300
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.1301 | 0.05 | 20 | 0.1112 |
| 0.14 | 0.09 | 40 | 0.1021 |
| 0.1582 | 0.14 | 60 | 0.1935 |
| 0.1106 | 0.19 | 80 | 0.1298 |
| 0.1646 | 0.23 | 100 | 0.1226 |
| 0.0748 | 0.28 | 120 | 0.0889 |
| 0.1169 | 0.33 | 140 | 0.0966 |
| 0.1127 | 0.38 | 160 | 0.0706 |
| 0.0899 | 0.42 | 180 | 0.0696 |
| 0.1202 | 0.47 | 200 | 0.1355 |
| 0.0491 | 0.52 | 220 | 0.0529 |
| 0.0931 | 0.56 | 240 | 0.0466 |
| 0.1068 | 0.61 | 260 | 0.0695 |
| 0.0545 | 0.66 | 280 | 0.0383 |
| 0.0346 | 0.7 | 300 | 0.0307 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.16.1
- Tokenizers 0.15.0