Mtplx
favorisiert
Anwendung
3x faster speeds on MLX | Qwen 3.8 27B | Native MTP Speculative Decoding On Apple Silicon With No External Drafter.
Aktiv gepflegte Filament-Plugins, Laravel-Pakete und Open-Source-Starter-Kits. Alle MIT, mit Releases für v3, v4 und v5, sofern zutreffend.
3x faster speeds on MLX | Qwen 3.8 27B | Native MTP Speculative Decoding On Apple Silicon With No External Drafter.
Giant MoE models on a single consumer GPU by streaming experts from SSD. CUDA fork of antirez/ds4: runs GLM-5.2 (743B), Tencent Hy3 (295B), and DeepSeek 4 Flash, with io_uring expert streaming, LFU host cache, cross-layer expert prefetch, and the first MTP speculative decoding for GLM-5.2.
Alle Pakete haben CI, Pest-Tests und offene Issues mit dem Tag good first issue.
Neue Version verfügbar.