Stars
Electron + Vite + React + Sass boilerplate.
verl: Volcano Engine Reinforcement Learning for LLMs
Undetected version of the Playwright testing and automation library.
A working pattern for SSE-based MCP clients and servers
binary-husky / trl
Forked from huggingface/trlTrain transformer language models with reinforcement learning.
🌐 Make websites accessible for AI agents. Automate tasks online with ease.
A Flexible Framework for Experiencing Cutting-edge LLM Inference Optimizations
A small utility library that enables you to execute Next.js server actions in parallel - the missing ingredient to build a boilerplate-free tRPC-style server-actions workflow.
A generative world for general-purpose robotics & embodied AI learning.
Xiaomi Home Integration for Home Assistant
A scriptable music downloader for Qobuz, Tidal, SoundCloud, and Deezer
The Postgres development platform. Supabase gives you a dedicated Postgres database to build your web, mobile, and AI applications.
Train transformer language models with reinforcement learning.
Use PEFT or Full-parameter to CPT/SFT/DPO/GRPO 500+ LLMs (Qwen3, Qwen3-MoE, Llama4, InternLM3, DeepSeek-R1, ...) and 200+ MLLMs (Qwen2.5-VL, Qwen2.5-Omni, Qwen2-Audio, Ovis2, InternVL3, Llava, GLM4…
Qlib is an AI-oriented Quant investment platform that aims to use AI tech to empower Quant Research, from exploring ideas to implementing productions. Qlib supports diverse ML modeling paradigms, i…
Production-ready platform for agentic workflow development.
Convert any URL to an LLM-friendly input with a simple prefix https://r.jina.ai/
This repo is a pipeline of VITS finetuning for fast speaker adaptation TTS, and many-to-many voice conversion
proxychains ng (new generation) - a preloader which hooks calls to sockets in dynamically linked programs and redirects it through one or more socks/http proxies. continuation of the unmaintained p…
pycorrector is a toolkit for text error correction. 文本纠错，实现了Kenlm，T5，MacBERT，ChatGLM3，Qwen2.5等模型应用在纠错场景，开箱即用。