Open-source ChatGPT / Comprehensive / Lightweight / Easy-to-use
The code is open-source, feel free to use it, contributions are welcome! Note: The license of the model depends on the provider of the model.
- 💥Latest News
- 💫OpenLLaMA2
- 💫Development Plan
- ⛏️Usage Steps
- 📄Running Example
- 📄Result Display
- 💐References & Acknowledgements
- 🌟Sponsor Us
- 🌈Starchart
- 🏆Contributors
-
2023/8/26: Support wandb logs
-
2023/8/22: Support ceval
-
2023/8/20: Add some PPO vs SFT examples
-
2023/8/18: support LLaMA2 7B PPO training on Single A100
pretraind SFT/RM checkpoint: https://huggingface.co/chuyi777/openllama2_checkpoint
-
2023/8/13: LLaMA2 7B + SFT+ RM + PPO + DeepSpeed training features finished
-
2023/07/30: OpenLLaMA2 project officially launched:
- Initial code submission
OpenLLaMA2 aims to develop a high-performance distributed LLaMA2 SFT/RLHF training framework.
The sister project of this project is chinese-llama2 ↗, which aims to fine-tune the Chinese LLaMA2 using SFT/RLHF.
- [✔️] Develop a fast LLaMA2 SFT/PPO Training Framework based on DeepSpeed.
- [✔️] Develop the Multi-nodes training scripts for Slurm.
- [✔️] Add wandb log support.
- [✔️] Support conda env.
- [WIP] Develop Multi-nodes RLHF based on Ray.
- [WIP] Support Multiple RM models.
- [WIP] Develop the Rejection Sampling.
- [WIP] Support QLora.
- [WIP] Support FlashAttention.
- [WIP] Develop the DPO.
- [WIP] Develop the RLHF datasets ↗ for Multiple reward models.
- [WIP] Train a chinese-llama2 ↗ RLHF model.
- [TODO] Develop the Context Distillation.
- [TODO] Training/Inference kernel fusion (such as DS inference)
- [TODO] Large-scale model (> 70B) support with ZeRO++ and FasterTransformer inference.
- [TODO] Better docs and examples
Clone the repository: git clone https://github.com/openllmai/OpenLLaMA2.git
-
Verified envs You can build openllama2 from nvidia docker(recomended) or from conda envs.
- Python: 3.8/3.9/3.10
- Torch: 2.0.0/2.0.1
- CUDA: 12.0+(recomended)/11.8/11.7
-
Single-node training with nvidia-docker
cd examples/scripts
# install nvidia-docker (Optional)
./nvidia_docker_install.sh
# launch nvidia container
./docker_run.sh
# cd in container
cd /openllama2/examples/scripts
# build OpenLLaMA2 (i.e, pip install)
./build_openllama2.sh
# huggingface login
~/.local/bin/huggingface-cli login
# train SFT model
./train_sft_llama.sh
# train RM model
./train_rm_llama.sh
# train PPO model
./train_ppo_llama.sh- Multi-nodes training on Slurm
cd examples/scripts
# huggingface login on Slurm
pip install transformers
huggingface-cli login
# Moidfy the Slurm Account/Nodes ... in `train_llama_slurm.sh`
# For SFT, RM, and PPO training stage:
# Modify the variable `training_script` in `train_llama_slurm.sh` to
readonly training_script="train_sft_llama.sh"
readonly training_script="train_rm_llama.sh"
readonly training_script="train_ppo_llama.sh"
# set `GPUS_PER_NODE` in `train_llama_slurm.sh`
readonly GPUS_PER_NODE=8
# run multi-nodes training script
# train_llama_slurm.sh will load the training args from `training_script`
sbatch ./train_llama_slurm.sh- build openllama2 from conda envs
If you really don't want to use nvidia docker, we also provide tutorials for building openllama2 from a conda environment
# we need conda
conda create -n llama2 python=3.8
# now requirements.txt seems incomplete for a conda env
pip install -r requirements.txt
# so, we need install some package manualy: when installing torch, you may need to match the corresponding cuda version.
pip install packaging ninja
pip install torch --index-url https://download.pytorch.org/whl/cu118
# check ninjia
ninja --version
echo $?
# install flash-attn: may take some time
pip install flash-attn --no-build-isolation
# build: may need to remove --use-feature=in-tree-build
./build_openllama2.sh
# enjoy it!After completing the training, you can evaluate your model by using the inference script:
./inference_llama.sh { model_path } "Please introduce the GTA5 game."We would like to express our gratitude to the following projects and organizations for their contributions to the field of AI and NLP:
How to Join?
- Email us at xianyuai@openllmai.top(official email) or janhu9527@gmail.com/jjgxw@outlook.com(PIC). Please include the following details:
- Your name
- Your GitHub username
- Your areas of interest
- Your skills and experience related to NLP and/or AI
- You can also join us through the official GitHub OpenLLaMA2 ↗ project page. Just create an issue about your interest to contribute and we will get back to you.
What can you do?
- Join the team and participate in the development of the OpenLLaMA2 project.
- Contribute to the project by submitting pull requests.
- Help improve documentation, fix bugs, or create new features.
- Share the project and help us grow the community.
Your sponsorship can help us maintain and improve OpenLLaMA2. If you find this project useful, please consider sponsoring us. You can sponsor us on Open Collective ↗.
A big thank you to all our contributors! If you want to contribute, feel free to make a pull request or create an issue.
@misc{openllmai23,
author = {OpenLLMAI},
title = {OpenLLaMA2},
year={2023},
howpublished = {\url{https://github.com/OpenLLMAI/OpenLLaMA2}}
}
OpenLLaMA2 © 2023 OpenLLMAI. All Rights Reserved.