Zhenyi Lu

Twin-Merging. [NeurIPS2024] Twin-Merging: Dynamic Integration of Modular Expertise in Model Merging

MIRACLE. [EMNLP2023]: MIRACLE: Towards Personalized Dialogue Generation with Latent-Space Multiple Personal Attribute Control

lm-evaluation-harness-fast. speedup for lm-evaluation-harness; support tensor-parallel inference and data-parallel inference; support gptq, bitsandbytes, peft and exllamav2.

Awesome-Model-Fusion. This GitHub repository summarizes most recent papers and resources related to the model fusion.

HELM-Extended-Local. support for various name in http/local run; support for gptq/bnb/tensorparallel in local run

CustomLLMFinetuningHandbook. an example to fine-tuning a Language Learning Model (LLM) from data preparation to deployment.

Boys-ToolKit. multi-process openai client; vllm data parallelism

Tools-gradio. custom Widget using gradio

Paper-Reading-.

axolotl. axolotl customize

dpo. Robust recipes for to align language models with human and AI preferences

LoRA-Pro-fix. Official code for our paper, "LoRA-Pro: Are Low-Rank Adapters Properly Optimized? "

trl. Train transformer language models with reinforcement learning.

PythonNotes. python notes for myself

peft. 🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.

Qwen-VL-DPO. Python

vllm. A high-throughput and memory-efficient inference and serving engine for LLMs

luogu-lzy. https://www.luogu.com.cn/ problems solving

Tools-for-HuggingfaceTransformers. custom tools