MobiusDevelopment/ Qwen3.6-35B-A3B-A8E4-128K:latest

19 yesterday

Mixed-precision quant of Qwen3.6-35B-A3B: 8-bit attention, 4-bit experts.

... /
params
87cd926f644a · 102B
{
"min_p": 0,
"num_batch": 2048,
"num_ctx": 131072,
"num_gpu": 999,
"temperature": 0.6,
"top_k": 20,
"top_p": 0.95
}