minimal-cost for training 0.5B R1-Zero

Ola: Pushing the Frontiers of Omni-Modal Language Model

MoBA: Mixture of Block Attention for Long-Context LLMs

Multi-lingual large voice generation model, providing inference, training and deployment full-stack ability.

A jounery to real multimodel R1 ! We are doing on large-scale experiment

A generative speech model for daily dialogue.

RLHF experiments on a single A100 40G GPU. Support PPO, GRPO, REINFORCE, RAFT, RLOO, ReMax, DeepSeek R1-Zero reproducing.

An Easy-to-use, Scalable and High-performance RLHF Framework (70B+ PPO Full Tuning & Iterative DPO & LoRA & RingAttention & RFT)

A fork to add multimodal model training to open-r1

Fully open reproduction of DeepSeek-R1

An Approach to Enhancing the Efficacy of Post-Training Using Synthetic Data by Iterative Data Selection

A series of technical report on Slow Thinking with LLM

Unified KV Cache Compression Methods for Auto-Regressive Models

LLM KV cache compression made easy

VideoLLaMA 2: Advancing Spatial-Temporal Modeling and Audio Understanding in Video-LLMs

Optimizing inference proxy for LLMs

OS-ATLAS: A Foundation Action Model For Generalist GUI Agents

Efficient Triton Kernels for LLM Training

Movie Gen Bench - two media generation evaluation benchmarks released with Meta Movie Gen

Inpaint anything using Segment Anything and inpainting models.

📖 Full Stack Practice of the Large Language Model Training @ RLChina 2024

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, I…