mistralit2_1000_STEPS_SFT_SFT
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: 0.2876
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-07
- train_batch_size: 4
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- training_steps: 1000
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.498 | 0.1 | 50 | 0.3952 |
| 0.3309 | 0.2 | 100 | 0.3213 |
| 0.3234 | 0.29 | 150 | 0.3104 |
| 0.2953 | 0.39 | 200 | 0.3048 |
| 0.2967 | 0.49 | 250 | 0.3005 |
| 0.3047 | 0.59 | 300 | 0.2972 |
| 0.2869 | 0.68 | 350 | 0.2943 |
| 0.2912 | 0.78 | 400 | 0.2913 |
| 0.2859 | 0.88 | 450 | 0.2895 |
| 0.2941 | 0.98 | 500 | 0.2880 |
| 0.2412 | 1.07 | 550 | 0.2886 |
| 0.2637 | 1.17 | 600 | 0.2884 |
| 0.2627 | 1.27 | 650 | 0.2882 |
| 0.2443 | 1.37 | 700 | 0.2881 |
| 0.2557 | 1.46 | 750 | 0.2877 |
| 0.259 | 1.56 | 800 | 0.2876 |
| 0.2598 | 1.66 | 850 | 0.2875 |
| 0.2633 | 1.76 | 900 | 0.2876 |
| 0.2727 | 1.86 | 950 | 0.2876 |
| 0.2674 | 1.95 | 1000 | 0.2876 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.0.0+cu117
- Datasets 2.18.0
- Tokenizers 0.15.2