Skip to content

mayafree-ai/Huggingface-MAYA

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🤗 Huggingface-MAYA — MAYA AI Spaces Mirror

Open source mirror of MAYA AI's flagship Hugging Face Spaces. Each subdirectory is a standalone, deployable Space — fork, clone, or deploy to your own HF account.

Hugging Face — MAYA-AI License Landing


📦 What's Inside

This repository contains 4 production-ready HF Spaces maintained by MAYA AI. Each folder is a complete Space — drop it into a new HF Space repo and it runs.

Folder HF Space SDK Likes Purpose
all-leaderboard/ MAYA-AI/all-leaderboard static ⭐ 28 Unified LLM leaderboard (CLIcK, KMMLU, GPQA, HAE-RAE, etc.)
qwen-3_5-chat/ MAYA-AI/QWEN-3_5-CHAT gradio ⭐ 16 Qwen 3.5 chat interface with HF OAuth
openclaw-moltbot/ mayafree/openclaw-moltbot gradio ⭐ 6 LLM battle arena (blind side-by-side voting)
fish-s2-pro-zero/ MAYA-AI/fish-s2-pro-zero gradio ⭐ 3 Zero-shot speech-to-speech voice cloning

🗂️ Repository Structure

Huggingface-MAYA/
├── README.md                  ← you are here
├── all-leaderboard/           ← 🏆 static HTML leaderboard
│   ├── README.md
│   ├── index.html             (40 KB — main leaderboard)
│   ├── backup.html
│   ├── style.css
│   └── .gitattributes
├── qwen-3_5-chat/             ← 💬 Qwen 3.5 chat (Gradio)
│   ├── README.md
│   ├── app.py                 (16 KB — Gradio app + OAuth)
│   ├── index.html             (47 KB — custom chat UI)
│   ├── requirements.txt
│   └── .gitattributes
├── openclaw-moltbot/          ← 🤖 LLM battle arena (Gradio)
│   ├── README.md
│   ├── app.py
│   ├── battle_arena.py        (31 KB — arena logic)
│   ├── index.html
│   ├── requirements.txt
│   └── .gitattributes
└── fish-s2-pro-zero/          ← 🗣️ voice cloning (Gradio)
    ├── README.md
    ├── app.py                 (12 KB — S2S inference)
    ├── requirements.txt
    ├── packages.txt
    └── .gitattributes

🚀 Deploy to Your Own HF Space

Each subdirectory is 1:1 compatible with a Hugging Face Space repository.

Option A — Create new Space on HF, push folder contents

# 1. Create a new Space on https://huggingface.co/new-space
#    (pick the same SDK: static / gradio / docker)

# 2. Clone it locally
git clone https://huggingface.co/spaces/YOUR_USERNAME/YOUR_SPACE
cd YOUR_SPACE

# 3. Copy this repo's folder contents into it
cp -r /path/to/Huggingface-MAYA/all-leaderboard/* .

# 4. Push
git add .
git commit -m "Initial deployment from MAYA AI mirror"
git push

Option B — Direct clone from HF (upstream)

git clone https://huggingface.co/spaces/MAYA-AI/all-leaderboard

🔐 Required Environment Variables (HF Space Secrets)

Some Spaces require secrets configured in the HF Space Settings → Variables and secrets panel.

qwen-3_5-chat/

Secret Required Purpose
HF_TOKEN ✅ yes Access gated Qwen models
OAUTH_CLIENT_ID ✅ yes HF OAuth login (auto-provided by HF Space OAuth)
OAUTH_CLIENT_SECRET ✅ yes HF OAuth login (auto-provided by HF Space OAuth)
SPACE_HOST optional Custom domain override (default: localhost:7860)
OAUTH_SCOPES optional OAuth scopes (default: openid profile)

openclaw-moltbot/

Secret Required Purpose
MAIN_CODE ✅ yes Runtime loader code (supplied by maintainer)

all-leaderboard/ · fish-s2-pro-zero/

No secrets required — runs out of the box.


🧬 Flagship Focus: all-leaderboard

Our leaderboard aggregates results across:

  • Korean benchmarks: CLIcK, KMMLU, HAE-RAE, KoBEST
  • Reasoning benchmarks: GPQA, MMLU, ARC-Challenge
  • Chat/Instruction: MT-Bench, AlpacaEval

Top-ranking MAYA AI models:

  • 🥇 Darwin-31B-Opus — CLIcK 84.5 / KMMLU 76.0 / GPQA 85.9
  • 🥇 Darwin-27B-Opus — GPQA 86.9 (SOTA in size class)

Explore live leaderboard


🔄 Sync Policy

This mirror is updated periodically. For the absolute latest version, always pull from the upstream HF Spaces — this GitHub mirror exists for:

  1. Discoverability — GitHub search surfaces code that HF doesn't index the same way
  2. Version history — commit-by-commit diffs across deployments
  3. Backup — independent snapshot of Space source
  4. Community contributions — PRs, issues, forks

🤝 Contributing

Found a bug? Want to improve a Space? PRs welcome.

  1. Fork this repo
  2. Make your changes in the relevant <space-name>/ folder
  3. Test locally (see each folder's README for run instructions)
  4. Open a PR — we'll review and mirror approved changes upstream to HF

📫 Contact


📜 License

This repository and its mirror contents are licensed under Apache License 2.0.

Individual Space components may depend on models or libraries with their own licenses — check each folder's README and the referenced model cards on Hugging Face.


Last synced: 2026-04-22 · Mirrored from huggingface.co/MAYA-AI and huggingface.co/mayafree

About

Open-source mirror of 4 flagship MAYA AI Hugging Face Spaces (all-leaderboard, QWEN-3_5-CHAT, openclaw-moltbot, fish-s2-pro-zero) ? each folder is a deployable Space

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors