最有价值项目

Free, open-source, offline, safe and secure AI Cantonese transcription, in your device.

Converts Text to Speech, Speech to Text, Support Streaming, Supports 4 Providers (OpenAI, Azure, ElevenLabs, Native Browser)

A project combining roguelike with LLMs, RAG, Text2Speech, and Speech2Text

📚 It's a Lecturer to you use when you will read some text

基于Dolphin模型的东方语言音视频转字幕api及webui

a simple speech recognition app using the Web Speech API Interfaces

🎙️ Uzbek system-wide speech-to-text dictation for macOS (rubaiSTT + whisper.cpp, Metal, offline). Press Ctrl+Opt+D anywhere.

Talk to your personal GPT Voice Assistant like Google Assistant or Alexa 🇪🇸🇬🇧, locally on your MacOS/Linux machine 🎤

A self-hosted voice-tracking teleprompter, based on https://voiceprompt.live but in Elixir Phoenix LiveView

Semi automated youtube channel that has a lot of cool features for someone to use in their content generating project

Multimodal AI Assistant is an advanced chatbot designed for local interactions with LLMs, files, and multimodal functionalities. Built on the Chainlit framework, it integrates Langchain and ChatOllama for hosting models locally, with cloud storage and monitoring via LiteralAI.

Archived (2026) — superseded by gbrlpzz/transcribe, a fully local macOS dictation & transcription tool. Scribe: Python transcription toolkit turning audio/video into titled PDFs via OpenAI Whisper + GPT.

🎤 Privacy-first local speech-to-text dictation for NixOS - Whisper.cpp powered push-to-talk with real-time feedback

A command line interface to combine text information from subtitles with voice data in the video. Provides a convenient way to generate training data for speech-recognition purposes.

ICE-PIXIU：A Cross-Language Financial Megamodeling Framework

Serve Poolside Laguna S 2.1 (NVFP4) on the NVIDIA DGX Spark (GB10) without hanging your box. Working stack, crash-safe configs, benchmarks, and the exact gotchas.

A curated list of tools, frameworks, and resources for running AI locally — inference, RAG, orchestration, monitoring, and more.

Intelligent load balancer for distributed vLLM server clusters 分布式 vLLM 服务器集群的智能负载均衡器

Layered prefill changes the scheduling axis from tokens to layers and removes redundant MoE weight reloads while keeping decode stall free. The result is lower TTFT, lower end-to-end latency, and lower energy per token without hurting TBT stability.

[AAAI 2025 oral] Attribution Analysis Meets Model Editing: Advancing Knowledge Correction in Vision Language Models with VisEdit