Wan: Open and Advanced Large-Scale Video Generative Models

Video-R1: Towards Super Reasoning Ability in Video Understanding MLLMs

Lighteval is your all-in-one toolkit for evaluating LLMs across multiple backends

Reproduce R1 Zero on Logic Puzzle

The Gaussian Histogram Loss (HL-Gauss) proposed by Imani et al. with a few convenient wrappers for regression, in Pytorch

Hunyuan-DiT : A Powerful Multi-Resolution Diffusion Transformer with Fine-Grained Chinese Understanding

Implementation of Rotary Embeddings, from the Roformer paper, in Pytorch

A fork to add multimodal model training to open-r1

s1: Simple test-time scaling

Official PyTorch Implementation of "Diffusion Autoencoders are Scalable Image Tokenizers"

PyTorch implementation of Advantage Actor Critic (A2C), Proximal Policy Optimization (PPO), Scalable trust-region method for deep reinforcement learning using Kronecker-factored approximation (ACKT…

AllenAI's post-training codebase

Arxiv个性化定制化模版，实现对特定领域的相关内容、作者与学术会议的有效跟进。

Quickly rewrite git repository history (filter-branch replacement)

Autoregressive Model Beats Diffusion: 🦙 Llama for Scalable Image Generation

Janus-Series: Unified Multimodal Understanding and Generation Models

Fully open reproduction of DeepSeek-R1

Clean, minimal, accessible reproduction of DeepSeek R1-Zero

starter from "How to Train a GAN?" at NIPS2016

This repo contains the code for 1D tokenizer and generator

Implementation of TiTok, proposed by Bytedance in "An Image is Worth 32 Tokens for Reconstruction and Generation"

Code for NeurIPS 2024 paper - The GAN is dead; long live the GAN! A Modern Baseline GAN - by Huang et al.

Using Low-rank adaptation to quickly fine-tune diffusion models.

Implementation of Mind Evolution, Evolving Deeper LLM Thinking, from Deepmind

A suite of image and video neural tokenizers

Unofficial implementation of Titans, SOTA memory for transformers, in Pytorch