metadata
base_model: meta-llama/Llama-2-7b-chat-hf
library_name: peft
license: llama2
tags:
- trl
- sft
- generated_from_trainer
model-index:
- name: content
results: []
content
This model is a fine-tuned version of meta-llama/Llama-2-7b-chat-hf on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.2418
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: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- 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: 0.05
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.594 | 0.4 | 2 | 2.4831 |
| 2.6396 | 0.8 | 4 | 2.3753 |
| 2.3316 | 1.2 | 6 | 2.2924 |
| 2.4193 | 1.6 | 8 | 2.2510 |
| 2.3204 | 2.0 | 10 | 2.2418 |
Framework versions
- PEFT 0.12.0
- Transformers 4.43.3
- Pytorch 2.1.2
- Datasets 2.16.0
- Tokenizers 0.19.1