Hert4/crab-agent-extension
Crab-Agent is an LLM-powered Chrome extension that automates browser tasks using natural language commands. Inspired by Claude for Chrome
Just doing my work
Repositories
Hert4/mochi-desktop
A local-first terminal pet companion that remembers you.
Hert4/ai-infra-notes
Ghi chép kỹ thuật về hạ tầng suy luận AI: phục vụ mô hình, huấn luyện, lượng tử hoá, đánh giá
Hert4/mstar
A high-performance, universal serving framework for any-to-any models.
Hert4/avabench
Agentic benchmark for MISA's AVA assistant, AutomationBench-style stateful world. Public 27-task set; leaderboard over the full 100.
Hert4/Qwen-Omni-Training-Talker
Omni talker training test
Hert4/yet-another-anime-game-launcher
Discord server https://discord.gg/HrV52MgSC2
Hert4/trace2skill
Harness-agnostic framework for evolving agent skills from real trajectories. Implementation of Trace2Skill: Distill Trajectory-Local Lessons into Transferable Agent Skills ([arXiv:2603.25158](https://arxiv.org/abs/2603.25158)).
Hert4/automationbench-opera-gemma4
AutomationBench 600 (Zapier, public set) results for open-weight models: Qwen3.6-27B base, beyoru/Opera, gemma-4-31B-it, beyoru/Ops-I. Per-task and per-assertion data + serving runbook.
Hert4/vllm-omni
A framework for efficient model inference with omni-modality models
Hert4/bitchat
bluetooth mesh chat, IRC vibes
Hert4/vllm
A high-throughput and memory-efficient inference and serving engine for LLMs
Hert4/sglang
SGLang is a high-performance serving framework for large language models and multimodal models.
Hert4/Confucius4-TTS-mlx
MLX port of Confucius4-TTS for fast on-device multilingual speech synthesis.
Hert4/container
A tool for creating and running Linux containers using lightweight virtual machines on a Mac. It is written in Swift, and optimized for Apple silicon.
Hert4/vlegal-bench
Hert4/llm-compressor
Transformers-compatible library for applying various compression algorithms to LLMs for optimized deployment with vLLM
Hert4/mlx-audio
A text-to-speech (TTS), speech-to-text (STT) and speech-to-speech (STS) library built on Apple's MLX framework, providing efficient speech analysis on Apple Silicon.
Hert4/clawdex
Using codex-pet in claude code that all
Hert4/guidellm
Evaluate and Enhance Your LLM Deployments for Real-World Inference Needs
Hert4/ideogram-deployment
Just run indeogram image generation model in different enviroment
Hert4/LLM-Certainty-Consistency
Backend-agnostic black-box hallucination & RAG-faithfulness detection for LLMs — Probabilistic Certainty & Consistency (arXiv:2601.02574). Works on MLX / OpenAI / vLLM via token logprobs; no model internals, no training.
Hert4/clawd-window
Cute pixel-art desktop pet for Windows. Climbs your app windows Shimeji-style, eats files into the Recycle Bin, and reacts to Claude Code sessions via a bundled plugin.
Hert4/oh-my-live2d
应用于浏览器环境且开箱即用的Live2D组件, 它支持所有版本的Live2D模型, 使用方式足够简单并且高可自定义, 可以快速为您的个人网站添加Live2D看板娘, 使您的个人网站变得更具有特色.
Hert4/research-notebooks
Hert4/Instantid-mobile
Hert4/ml-fastvlm-v2
Hert4/openclaw
Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞
Hert4/moshi
Moshi is a speech-text foundation model and full-duplex spoken dialogue framework. It uses Mimi, a state-of-the-art streaming neural audio codec.
Hert4/pocket-tts
A TTS that fits in your CPU (and pocket)