Post #3167546

OLLAMA is getting TurboQuant support. Specialized models can already use it. So can I stuff a 256k context into my 7900XTX with a 36b model? Going to find out soon! https://ollama.com/iliafed/qwen3.6turboquant