Hugging Face's logo

MJPansa
/
Qwen3.8-Flash-Next-MXFP4-FP8-R12

Image-Text-to-Text
Safetensors
vllm
qwen4_exp
qwen
qwen3.8
qwen4-exp
mxfp4
mxfp8
fp8
w4a8
mixed-precision
mixture-of-experts
vision-language
dgx-spark
conversational
Model card Files
xet
 Community
Qwen3.8-Flash-Next-MXFP4-FP8-R12
147 GB
  • 1 contributor
 History: 14 commits
MJPansa's picture
MJPansa
Reduce documented single-Spark KV cache to 700K slots
dca049e verified