metadata
library_name: peft
license: other
base_model: meta-llama/Llama-3.1-8B-Instruct
tags:
- llama-factory
- lora
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: factory_llama_results
results: []
factory_llama_results
This model is a fine-tuned version of meta-llama/Llama-3.1-8B-Instruct on the train dataset. It achieves the following results on the evaluation set:
- Loss: 0.2624
- Accuracy: 0.9526
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: 2
- eval_batch_size: 2
- seed: 42
- distributed_type: multi-GPU
- num_devices: 3
- gradient_accumulation_steps: 4
- total_train_batch_size: 24
- total_eval_batch_size: 6
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 9.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.3833 | 1.0 | 42 | 0.3712 | 0.9116 |
| 0.298 | 2.0 | 84 | 0.2805 | 0.9280 |
| 0.2038 | 3.0 | 126 | 0.2475 | 0.9400 |
| 0.1427 | 4.0 | 168 | 0.2243 | 0.9458 |
| 0.1081 | 5.0 | 210 | 0.2245 | 0.9490 |
| 0.066 | 6.0 | 252 | 0.2289 | 0.9516 |
| 0.0503 | 7.0 | 294 | 0.2457 | 0.9523 |
| 0.0401 | 8.0 | 336 | 0.2616 | 0.9527 |
| 0.0338 | 8.7904 | 369 | 0.2624 | 0.9526 |
Framework versions
- PEFT 0.15.2
- Transformers 4.51.3
- Pytorch 2.7.1+cu126
- Datasets 3.6.0
- Tokenizers 0.21.1