Taobao MNN-Pocket-0.3B trends on Hugging Face: Qwen3-based on-device INT4 model
taobao-mnn · hf · 2026-11-10
Taobao's MNN team released MNN-Pocket-0.3B, now trending on Hugging Face in text generation.
A 0.3B-parameter model distilled from Qwen3, quantized to INT4 for on-device and edge deployment, published in safetensors format and compatible with both transformers and the MNN ecosystem.
More from Infra
- How banned Nvidia chips actually reach China: stolen trucks, fake paperwork, hair-dryer sticker swaps — Caneki97 · 2026-11-10
- Tencent reportedly picks Bittensor subnet Chutes as launch partner for flagship model Hy4 — markjeffrey · 2026-11-10
- Cutting GPU power 42% costs nothing: RTX 3080 Ti runs Qwen 3.5 9B at full speed on 250W — Asthenia5 · 2026-11-10
- PyTorch's CUDA Tile: a Python DSL for tiling kernels onto Tensor Cores — PyTorch · 2026-11-10
- Indie dev has an AI agent migrate off Zeabur to Dokploy in 2 hours, cancels all servers — yihui_indie · 2026-11-10
- Local inference shootout: Basalt beats Strata at 118 tok/s on Qwen3.8 with half the RAM — x_MASE_x · 2026-11-10