llm

分享高质量的AI项目

Open Fiesta lets you chat with 100+ AI models like OpenAI, Gemini, Claude, Perplexity, Deepseek, and Grok in one place. Compare model responses side-by-side in real-time and choose the best AI for every task

A RAG system designed to process documents with multimodal content. It can generate factual, context-aware answers to user queries, based on the documents texts, tables, figures, ...

Use your X Premium or SuperGrok subscription in pi.

Access 100s of LLMs with minimal lines of code

Open-source, voice-first AI mock interviewer. Upload your CV + a job description, practice out loud with an adaptive multilingual interviewer, get scored and coached on your weak areas. LiveKit · LangGraph · Next.js

This project is the backend engine for a fully autonomous AI-powered call center. It integrates a large language model (LLM), speech recognition, and text-to-speech to manage real-time phone conversations via Asterisk.

Official implementation for Text Generation Beyond Discrete Token Sampling

Python SDK for Agent Vector Protocol – transfer KV-cache between LLM agents instead of text

Hands-on labs for Microsoft Foundry - Azure's unified PaaS for enterprise AI. Notebooks + Bicep covering provisioning, agents (incl. hosted Copilot SDK + REST), MCP, Foundry IQ knowledge bases, guardrails, red-teaming, and fine-tuning.

An LLM training library for instruction-tuning.

Local LLM chat app for Apple Silicon Macs.

Production-ready multi-agent platform delivering orchestration, session governance and audit capabilities. Stack: Go/Python/Rust microservices, multi-type data storage, Next.js admin UI, automated CI/CD pipelines, AI code review triggered via PR comment mentions.

Enterprise-grade browser extension bringing multilingual voice interaction to AI chatbots (Pi, Claude, ChatGPT). Features real-time speech detection with Silero VAD, accurate transcription via OpenAI Whisper, and ElevenLabs TTS. Built with TypeScript, XState, and modern web standards. Progressive enhancement across Chrome, Firefox, and Safari

Probing the limitations of multimodal language models for chemistry and materials research

Training-free KV cache compression via E8 lattice VQ. 2-bit KV that preserves retrieval (30/30 NIAH vs TurboQuant 0/30). Calibration-free, 9 architectures validated.

Unified KV-cache compression for LLM inference: 12 Python-native methods, guarded add-on composition and routing, analytical capacity simulation, Godzilla KVarN/TriAttention with SM86/SM89 qualification, CUDA weight sharing, and multi-GPU planning.

Kubernetes scanner that discovers LLMs running on vLLM and extracts their deployment and runtime facts.

Awesome-HCI （Ubiquitous, LLM, MLLM, Agent, RAG, Embodied-AI, RLHF)

Local AI workstation — discover, run, chat, benchmark, and generate images from open-weight models. DFlash/DDTree speculative decoding, TurboQuant & TriAttention cache compression strategies, MLX + llama.cpp + vLLM + MTPLX backends.