metadata
license: llama2
base_model: meta-llama/Llama-2-7b-hf
tags:
- alignment-handbook
- trl
- sft
- generated_from_trainer
- trl
- sft
- generated_from_trainer
datasets:
- HuggingFaceH4/ultrafeedback_binarized
model-index:
- name: llama2-7b-sft-full-llama2
results: []
llama2-7b-sft-full-llama2
This model is a fine-tuned version of meta-llama/Llama-2-7b-hf on the HuggingFaceH4/ultrafeedback_binarized dataset. It achieves the following results on the evaluation set:
- Loss: 1.0380
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: 2e-05
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 8
- total_train_batch_size: 128
- total_eval_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: 3
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.0183 | 0.9980 | 126 | 1.0127 |
| 0.869 | 1.9960 | 252 | 1.0026 |
| 0.714 | 2.9941 | 378 | 1.0380 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.3.0a0+ebedce2
- Datasets 2.19.2
- Tokenizers 0.19.1