Popular repositories Loading

1. maya-s24-turbo maya-s24-turbo Public

   GLM-5.3-Flash (Maya-S24) tuned on 2×RTX 3090: 59 t/s thinking-on @ 262k ctx — full tuning log, measured reject list, and benchkit (the model benchmarked itself). Built on mw00/project-maya

   HTML 11 2

2. Qwen3.8-Flash-Next-Single-DGX-Spark Qwen3.8-Flash-Next-Single-DGX-Spark Public

   Forked from MiaAI-Lab/Qwen3.8-Flash-Next-Single-DGX-Spark

   Qwen3.8-Flash-Next on ONE DGX Spark (TP=1)

   Python 3

3. Dspine-senorita Dspine-senorita Public

   DSpine speculative decoding on RTX 3090: measured speed campaign for Qwen3.8-27B EXL3 4bpw — dynamic verify, n-gram hybrid, drafter re-quant, OpenAI-compatible server (built on exllamav3)

   Python 1

4. Sato Sato Public

   A lean chat-only web app forked from deepseek-ai/deepseek-harness — minimal context, provider profiles, Chrome-bridge search, MIT

   TypeScript

5. qwen3090-turbo qwen3090-turbo Public

   Qwen3.8-27B W4A16 + DFlash2 on 2×RTX 3090: 256k ctx, 292k GPU KV + 48GB DRAM offload tier, zero quality loss — full tuning log & benchmarks

   Python

6. gemma-jev gemma-jev Public

   Local Gemma 4 E4B Jev alternative: ~33ms typed decisions on llama.cpp driving a visible Chrome web controller for Sato

   Python