superbonyx/gemma-4-26B-A4B-it-qat-assistant-MTP-Q8_0.gguf

Reorder card (Benchmark up top, Why below Conversion); Requirements: add target QAT GGUF

Usage: add full llama-server argument list used for the benchmark

Add model card (provenance, usage, benchmark; fix license)

Add Q8_0 MTP draft head (gemma4-assistant, master-converted)

initial commit