Hugging Face's logo
Anurag's picture
šŸ¤ Open to Collab

Anurag

edwixx
26 40 153
Jitheshdr123's profile picture dvilasuero's profile picture AMAImedia's profile picture
Ā·
https://anuragkanade.com

AI & ML interests

ASR, TTS, Speech and Audio

Recent Activity

updated a dataset 2 days ago
edwixx/s2s-blog-assets
liked a dataset 2 days ago
hiraki/candor-turntaking-annotations
reacted to SeaWolf-AI's post with šŸ”„ 4 days ago
šŸ“± POCKET ā€” a 35-billion-parameter model that runs on your iPhone, and on your PC with no GPU We're releasing POCKET, VIDRAFT's flagship Darwin-36B-Opus compressed for on-device use. No fork, no CUDA, no cloud ā€” it runs on stock llama.cpp. It's a sparse Mixture-of-Experts model (256 experts, only 8 active per token), so the file can be large while the work per token stays small. That's what lets a 35B model run on a phone, and generate fast on a CPU with no graphics card. Measured (POCKET-35B IQ1_M vs Bonsai-27B Q1_0): ā€¢ CPU generate (Xeon, 16 threads): 27.0 vs 10.1 tok/s ā†’ 2.69Ɨ faster ā€¢ GPU generate (H100): 197 vs 89 tok/s ā†’ 2.22Ɨ faster ā€¢ GPU prompt processing (H100): 753 vs 1816 ā†’ 0.41Ɨ (Bonsai wins this one ā€” MoE prefill wakes every expert, so sparsity stops helping there. We say so.) ā€¢ Quality (HellaSwag, 400 q): 61.0% vs 60.0% ā†’ a tie (confidence intervals overlap) On a real consumer laptop ā€” MacBook M3 Pro (18 GB) ā€” POCKET wins every axis, prompt processing included: ā€¢ Metal generate: 25.4 vs 12.8 ā†’ 1.99Ɨ ā€¢ CPU generate: 13.8 vs 4.4 ā†’ 3.13Ɨ ā€¢ Metal prompt: 240.7 vs 73.4 ā†’ 3.28Ɨ One more quiet fact: the same-size, quality-oriented rival Ternary-Bonsai-27B (7.2 GB) fails to load in upstream llama.cpp at all ā€” it needs the PrismML fork. POCKET runs on the tools you already have: LM Studio, Ollama, PocketPal, MLX. šŸ“– Full story (tech, measurements, recipes): https://huggingface.co/blog/FINAL-Bench/pocket Models: šŸ“¦ POCKET-35B-GGUF (PC / server, no GPU): https://huggingface.co/FINAL-Bench/POCKET-35B-GGUF šŸ‡°šŸ‡· POCKET-KR-GGUF (Android): https://huggingface.co/FINAL-Bench/POCKET-KR-GGUF šŸŽ POCKET-KR-MLX (iPhone / Mac): https://huggingface.co/FINAL-Bench/POCKET-KR-MLX šŸŒ POCKET-EN-GGUF (English phone / PC): https://huggingface.co/FINAL-Bench/POCKET-EN-GGUF šŸ–„ļø Live demo (answering on a CPU, no GPU): https://huggingface.co/spaces/FINAL-Bench/POCKET-35B-CPU šŸ“š Collection: https://huggingface.co/collections/FINAL-Bench/pocket-models-6a618ee5d23eafb7e185a5c6
View all activity

Organizations

Stable Diffusion API's profile picture Tune a video concepts library's profile picture ModelsLab's profile picture StyleTTS 2 Community's profile picture MLX Community's profile picture Chinese LLMs on Hugging Face's profile picture Hugging Face Context Course's profile picture ICML 2026 Agent Reproductions's profile picture