Pretraining LLM from scratch

Use PEFT or Full-parameter to finetune 500+ LLMs (Qwen2.5, InternLM3, GLM4, Llama3.3, Mistral, Yi1.5, DeepSeek-R1, ...) and 200+ MLLMs (Qwen2.5-VL, Qwen2.5-Omni, Qwen2-Audio, Llama3.2-Vision, Llava…

Model interpretability and understanding for PyTorch

A demo to apply simultaneous interpretation based on Azure

Reproduction of paper "Lateral interaction by Lapalcian-based graph smoothing for deep neural networks"

Reproduce R1 Zero on Logic Puzzle

Official Repo for Open-Reasoner-Zero

Democratizing Reinforcement Learning for LLMs

One-click start reproduction of multi-modal DeepSeek R1-Zero

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

Integrate the DeepSeek API into popular softwares

Biomedical Question Answering Datasets.

Fully open reproduction of DeepSeek-R1

Clean, minimal, accessible reproduction of DeepSeek R1-Zero

The all-in-one Desktop & Docker AI application with built-in RAG, AI agents, No-code agent builder, and more.

RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine based on deep document understanding.

Agent Laboratory is an end-to-end autonomous research workflow meant to assist you as the human researcher toward implementing your research ideas

AllenAI's post-training codebase

Reproducible, flexible LLM evaluations

Data and tools for generating and inspecting OLMo pre-training data.

prime is a framework for efficient, globally distributed training of AI models over the internet.

仅需Python基础，从0构建大语言模型；从0逐步构建GLM4\Llama3\RWKV6， 深入理解大模型原理

Modeling, training, eval, and inference code for OLMo

A generative world for general-purpose robotics & embodied AI learning.

Official code for Coupled Oscillatory RNN (ICLR 2021, Oral)

phy: interactive visualization and manual spike sorting of large-scale ephys data

Fast spike sorting with drift correction

[NeurIPS 2019] Code for the paper "Learning to Control Self-Assembling Morphologies: A Study of Generalization via Modularity"