metadata
base_model: mistralai/Mistral-7B-Instruct-v0.3
datasets:
- GaetanMichelet/chat-60_ft_task-2
library_name: peft
license: apache-2.0
tags:
- alignment-handbook
- trl
- sft
- generated_from_trainer
model-index:
- name: Mistral-7B_task-2_60-samples_config-2_full
results: []
Mistral-7B_task-2_60-samples_config-2_full
This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.3 on the GaetanMichelet/chat-60_ft_task-2 dataset. It achieves the following results on the evaluation set:
- Loss: 0.8033
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 |
|---|---|---|---|
| 1.1698 | 0.6957 | 2 | 1.1609 |
| 1.1595 | 1.7391 | 5 | 1.0642 |
| 1.0152 | 2.7826 | 8 | 0.9874 |
| 0.9495 | 3.8261 | 11 | 0.9347 |
| 0.8878 | 4.8696 | 14 | 0.8748 |
| 0.823 | 5.9130 | 17 | 0.8343 |
| 0.7669 | 6.9565 | 20 | 0.8200 |
| 0.7616 | 8.0 | 23 | 0.8117 |
| 0.7438 | 8.6957 | 25 | 0.8073 |
| 0.7119 | 9.7391 | 28 | 0.8034 |
| 0.7076 | 10.7826 | 31 | 0.8033 |
| 0.6789 | 11.8261 | 34 | 0.8046 |
| 0.6409 | 12.8696 | 37 | 0.8047 |
| 0.625 | 13.9130 | 40 | 0.8139 |
| 0.581 | 14.9565 | 43 | 0.8231 |
| 0.5582 | 16.0 | 46 | 0.8310 |
| 0.5045 | 16.6957 | 48 | 0.8541 |
| 0.507 | 17.7391 | 51 | 0.8568 |
Framework versions
- PEFT 0.12.0
- Transformers 4.44.0
- Pytorch 2.1.2+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1