Lompat ke konten utama
buildradar
Sign in

carloslfu/slotstream

@carloslfu

Run Qwen3.8-Flash-Next (125B MoE, 104 GB at 4-bit) on Macs with a fraction of that RAM by streaming experts from SSD. MLX + Swift, Ollama-compatible API.

Bintang
287
Fork
16
Bahasa
Swift
Lisensi
MIT
Push terakhir
2 hari yang lalu
Swiftmacosllmswiftollamallm-inferencelocal-llmqwenapple-siliconmlxmixture-of-experts