HunyuanVideo-I2V: A Customizable Image-to-Video Model based on HunyuanVideo

We systematically studied the influencing factors when LLM generates benchmarks,By using our code, you can generate high-quality QA datasets

adds Sequence Parallelism into LLaMA-Factory

LoRA training extention for Stable Diffusion Web-UI

Dead simple FLUX LoRA training UI with LOW VRAM support

CogView4, CogView3-Plus and CogView3(ECCV 2024)

all you need to start with Le Chat Mistral API

LOTUS: A semantic query engine for fast and easy LLM-powered data processing

Data management with LLMs

Building DeepSeek R1 from Scratch

✨ AI interface for tinkerers (Ollama, Haystack RAG, Python)

🚀 DeepSeek-V3 R1大模型逆向API【特长：良心厂商】（官方贼便宜，建议直接走官方），支持高速流式输出、多轮对话，联网搜索，R1深度思考，零配置部署，多路token支持，仅供测试，如需商用请前往官方开放平台。

Solve Visual Understanding with Reinforced VLMs

wdtagger-lora

Forces DeepSeek R1 models to engage in extended reasoning by intercepting early termination tokens.

Multimodal Representation Alignment for Image Generation: Text-Image Interleaved Control Is Easier Than You Think!

MoD Control Tile Upscaler for SDXL Pipeline

Code for RealmDreamer: Text-Driven 3D Scene Generation with Inpainting and Depth Diffusion [3DV 2025]

Various AI scripts. Mostly Stable Diffusion stuff.

gui

GUI-focused roop

HunyuanVideo Keyframe Control Lora is an adapter for HunyuanVideo T2V model for keyframe-based video generation

Crawler and cleaner of data for novelai embedding's training

Test-time compute in information retrieval

Fast State-of-the-Art Static Embeddings

Local RAG researcher agent built using Langgraph, DeepSeek R1 and Ollama