metadata
base_model: mistralai/Mistral-7B-Instruct-v0.3
datasets:
- GaetanMichelet/chat-60_ft_task-1_auto
library_name: peft
license: apache-2.0
tags:
- alignment-handbook
- trl
- sft
- generated_from_trainer
model-index:
- name: Mistral-7B_task-1_60-samples_config-2_auto
results: []
Mistral-7B_task-1_60-samples_config-2_auto
This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.3 on the GaetanMichelet/chat-60_ft_task-1_auto dataset. It achieves the following results on the evaluation set:
- Loss: 0.7918
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: 1
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- gradient_accumulation_steps: 16
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.0805 | 0.6957 | 2 | 1.0819 |
| 1.0214 | 1.7391 | 5 | 0.9261 |
| 0.7668 | 2.7826 | 8 | 0.8322 |
| 0.7312 | 3.8261 | 11 | 0.7920 |
| 0.5041 | 4.8696 | 14 | 0.7918 |
| 0.4287 | 5.9130 | 17 | 0.8765 |
| 0.2628 | 6.9565 | 20 | 0.9647 |
| 0.1707 | 8.0 | 23 | 1.1113 |
| 0.1233 | 8.6957 | 25 | 1.2851 |
| 0.0474 | 9.7391 | 28 | 1.4284 |
| 0.0448 | 10.7826 | 31 | 1.4882 |
| 0.0198 | 11.8261 | 34 | 1.6129 |
Framework versions
- PEFT 0.12.0
- Transformers 4.44.0
- Pytorch 2.1.2+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1