georgesung's picture
Create README.md
dd775f6
|
raw
history blame
894 Bytes
metadata
license: other
datasets:
  - georgesung/OpenOrca_35k

Overview

Fine-tuned Llama-2 7B with a 35k subset of the OpenOrca dataset georgesung/OpenOrca_35k. Used QLoRA for fine-tuning. Trained for one epoch on a 24GB GPU (NVIDIA A10G) instance.

Prompt style

The model was trained with the following prompt style:

### System:
You are a helpful AI assistant.

### Instruction:
Hello

### Response:
Hi, how can I help you?

Training code

Code used to train the model is available here.

To reproduce the results:

git clone https://github.com/georgesung/llm_qlora
cd llm_qlora
pip install -r requirements.txt
python train.py configs/llama2_7b_orca_35k.yaml

Fine-tuning guide

https://georgesung.github.io/ai/qlora-ift/