Can a Sapphire PULSE run Qwen2.5-Coder 32B?
Sort of — Runs with compromises
Too big for 16 GB of VRAM (needs ~23 GB), but it can run by offloading part of the model to system RAM — slower, and it uses your RAM. A smaller quantization may help.
Your VRAM
16 GB
Needs (est.)
23 GB
Quantization
Q4_K_M