Creator

Victor Cruz

I quantize frontier models down into EXL3 and GGUF to fit on smaller GPUs, while also creating the most optimal recipes 2 entries for Victor Cruz from 1 creator on GitHub, published Sep 7 to Sep 17.

Qwen3.8-Flash-Next2

vcruz305

@vcruz305 · Sep 7

Serves turboderp's EXL3 quant of the model on one Spark using exllamav3.

vcruz305

@vcruz305 · Sep 17

Ports ExLlamaV3 kernels to RDNA 3.5 with grouped MoE, int8 mixers and MTP. The author quantizes frontier models down to fit smaller GPUs.

Entries