A fork to add multimodal model training to open-r1

Data and code for NeurIPS 2022 Paper "Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering".

This repository is build in association with our position paper on "Multimodality for NLP-Centered Applications: Resources, Advances and Frontiers". As a part of this release we share the informati…

A lightweight data processing framework built on DuckDB and 3FS.

A high-performance distributed file system designed to address the challenges of AI training and inference workloads.

A bidirectional pipeline parallelism algorithm for computation-communication overlap in V3/R1 training.

Official codebase for "SWE-RL: Advancing LLM Reasoning via Reinforcement Learning on Open Software Evolution"

DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling

EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL

FlashMLA: Efficient MLA decoding kernels

Solve Visual Understanding with Reinforced VLMs

Qwen2.5-VL is the multimodal large language model series developed by Qwen team, Alibaba Cloud.

This repo contains the code for "MEGA-Bench Scaling Multimodal Evaluation to over 500 Real-World Tasks" [ICLR2025]

verl: Volcano Engine Reinforcement Learning for LLMs

Sky-T1: Train your own O1 preview model within $450

LIMO: Less is More for Reasoning

s1: Simple test-time scaling

Clean, minimal, accessible reproduction of DeepSeek R1-Zero

Witness the aha moment of VLM with less than $3.

DeepSeek Coder: Let the Code Write Itself

Janus-Series: Unified Multimodal Understanding and Generation Models

Collect every awesome work about r1!

Fully open data curation for reasoning models

Fully open reproduction of DeepSeek-R1

[ACL 2024]Official GitHub repo for OlympiadBench: A Challenging Benchmark for Promoting AGI with Olympiad-Level Bilingual Multimodal Scientific Problems.

ModelScope

OpenCompass is an LLM evaluation platform, supporting a wide range of models (Llama3, Mistral, InternLM2,GPT-4,LLaMa2, Qwen,GLM, Claude, etc) over 100+ datasets.