Hugging Face's logo
Activity Feed

AI & ML interests

On-device AI, GGUF quantization, Apple Silicon, macOS automation

Recent Activity
hero775Β  updated a model 2 days ago
batiai/Kimi-K2.7-Code-GGUF
hero775Β  updated a model 3 days ago
batiai/batisay-ko-large
hero775Β  updated a model 3 days ago
batiai/batisay-ko-base
View all activity

batiai 's collections 7

πŸŽ™οΈ ν•œκ΅­μ–΄ μŒμ„± μŠ€μœ„νŠΈ β€” STT + ν™”μžλΆ„λ¦¬
batisay(STT, 무엇을 λ§ν–ˆλ‚˜) + batispeak(ν™”μžλΆ„λ¦¬, λˆ„κ°€ λ§ν–ˆλ‚˜) = ν†΅ν™”Β·νšŒμ˜ ν™”μžλ³„ 전사. 16GB Mac on-device.
πŸš€ Frontier MoE β€” 128B–1T
Largest open-weight LLMs, BatiAI-quantized. Mac-runnable from M4 Max 128GB to Mac Studio M3 Ultra 512GB.
🍎 Gemma 4 β€” Google's Latest
Gemma 4 quantizations from Google's official weights. Best entry for 16GB Mac mini M4 (E4B Q4 = 57 t/s).
BatiAI RAG Stack
Complete Mac-first on-device RAG stack β€” chat LLM + reranker + text/VL embedder, direct from BF16, BatiAI-signed. For BatiFlow.
🧠 NVIDIA Nemotron 3 β€” Hybrid Mamba+Attention
NVIDIA Nemotron 3 family β€” NemotronH architecture combining Mamba state-space + standard attention. Mac-runnable, BatiAI-quantized + signed.
πŸŽ™οΈ ν•œκ΅­μ–΄ μŒμ„± μŠ€μœ„νŠΈ β€” STT + ν™”μžλΆ„λ¦¬
batisay(STT, 무엇을 λ§ν–ˆλ‚˜) + batispeak(ν™”μžλΆ„λ¦¬, λˆ„κ°€ λ§ν–ˆλ‚˜) = ν†΅ν™”Β·νšŒμ˜ ν™”μžλ³„ 전사. 16GB Mac on-device.
🧠 NVIDIA Nemotron 3 β€” Hybrid Mamba+Attention
NVIDIA Nemotron 3 family β€” NemotronH architecture combining Mamba state-space + standard attention. Mac-runnable, BatiAI-quantized + signed.
πŸš€ Frontier MoE β€” 128B–1T
Largest open-weight LLMs, BatiAI-quantized. Mac-runnable from M4 Max 128GB to Mac Studio M3 Ultra 512GB.
🍎 Gemma 4 β€” Google's Latest
Gemma 4 quantizations from Google's official weights. Best entry for 16GB Mac mini M4 (E4B Q4 = 57 t/s).
BatiAI RAG Stack
Complete Mac-first on-device RAG stack β€” chat LLM + reranker + text/VL embedder, direct from BF16, BatiAI-signed. For BatiFlow.