Hugging Face's logo

autotrust
/
GLM5.3-Flash-E256-DGX-Spark

Image-Text-to-Text
Safetensors
English
Chinese
vllm
glm5_next
autotrust
Mixture of Experts
nvfp4
modelopt
dgx-spark
gb10
blackwell
neural-architecture-search
mtp
speculative-decoding
conversational
8-bit precision
Model card Files
xet
 Community
GLM5.3-Flash-E256-DGX-Spark / scripts /export_mtp_nvfp4.py

Commit History

30 days ago

  1. Fix mtp-nvfp4: re-export with ModelOpt (shared expert NVFP4), works on vLLM 0.31

    cloudyu
    Verified signature
    8045270