verl-ascend

Use PEFT or Full-parameter to finetune 450+ LLMs (Qwen2.5, InternLM3, GLM4, Llama3.3, Mistral, Yi1.5, Baichuan2, DeepSeek-R1, ...) and 150+ MLLMs (Qwen2.5-VL, Qwen2-Audio, Llama3.2-Vision, Llava, InternVL2.5, MiniCPM-V-2.6, GLM4v, Xcomposer2.5, Yi-VL, DeepSeek-VL2, Phi3.5-Vision, GOT-OCR2, ...).

Align Anything: Training All-modality Model with Feedback

MedicalGPT: Training Your Own Medical GPT Model with ChatGPT Training Pipeline. 训练医疗大模型，实现了包括增量预训练(PT)、有监督微调(SFT)、RLHF、DPO、ORPO、GRPO。

verl-agent is an extension of veRL, designed for training LLM/VLM agents via RL. verl-agent is also the official code for paper "Group-in-Group Policy Optimization for LLM Agent Training"

MixGRPO: Unlocking Flow-based GRPO Efficiency with Mixed ODE-SDE

Official code of DeepMesh: Auto-Regressive Artist-mesh Creation with Reinforcement Learning

Explore the Multimodal “Aha Moment” on 2B Model

tensorflow? ???? ??? ?????, Topic Models, BERT, GPT, LLM? ?? ?? ??? ????? ????? ??? Deep Learning NLP ??????.

Collect every awesome work about r1!

An Efficient "Factory" to Build Multiple LoRA Adapters