metadata
base_model: meta-llama/Meta-Llama-3-8B
datasets:
- generator
library_name: peft
license: llama3
tags:
- trl
- sft
- generated_from_trainer
model-index:
- name: Llama3_no_extraction_V2
results: []
Llama3_no_extraction_V2
This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 1.3288
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-05
- train_batch_size: 2
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- total_eval_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| No log | 0.9903 | 51 | 1.3462 |
| No log | 2.0 | 103 | 1.3212 |
| No log | 2.9903 | 154 | 1.3157 |
| No log | 4.0 | 206 | 1.3153 |
| No log | 4.9903 | 257 | 1.3163 |
| No log | 6.0 | 309 | 1.3196 |
| No log | 6.9903 | 360 | 1.3231 |
| No log | 8.0 | 412 | 1.3249 |
| No log | 8.9903 | 463 | 1.3273 |
| 1.1964 | 9.9029 | 510 | 1.3288 |
Framework versions
- PEFT 0.11.1
- Transformers 4.41.2
- Pytorch 2.3.1+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1