metadata
base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
datasets:
- GaetanMichelet/chat-60_ft_task-3_auto
library_name: peft
license: llama3.1
tags:
- alignment-handbook
- trl
- sft
- generated_from_trainer
model-index:
- name: Llama-31-8B_task-3_60-samples_config-2_auto
results: []
Llama-31-8B_task-3_60-samples_config-2_auto
This model is a fine-tuned version of meta-llama/Meta-Llama-3.1-8B-Instruct on the GaetanMichelet/chat-60_ft_task-3_auto dataset. It achieves the following results on the evaluation set:
- Loss: 0.3568
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 |
|---|---|---|---|
| 2.2034 | 0.6957 | 2 | 2.2399 |
| 2.1509 | 1.7391 | 5 | 1.8617 |
| 1.5862 | 2.7826 | 8 | 1.1118 |
| 0.9806 | 3.8261 | 11 | 0.5886 |
| 0.4601 | 4.8696 | 14 | 0.4766 |
| 0.4788 | 5.9130 | 17 | 0.4128 |
| 0.3828 | 6.9565 | 20 | 0.3872 |
| 0.2646 | 8.0 | 23 | 0.3888 |
| 0.349 | 8.6957 | 25 | 0.3620 |
| 0.299 | 9.7391 | 28 | 0.3568 |
| 0.2883 | 10.7826 | 31 | 0.3583 |
| 0.2187 | 11.8261 | 34 | 0.3658 |
| 0.2439 | 12.8696 | 37 | 0.3657 |
| 0.1651 | 13.9130 | 40 | 0.3713 |
| 0.2365 | 14.9565 | 43 | 0.3811 |
| 0.1654 | 16.0 | 46 | 0.3851 |
| 0.161 | 16.6957 | 48 | 0.4000 |
Framework versions
- PEFT 0.12.0
- Transformers 4.44.0
- Pytorch 2.1.2+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1