nvidia
/

Gemma-4-31B-IT-NVFP4

Text Generation

Model Optimizer

Model card Files Files and versions

Resources

View closed (2)

Why not quantize the MATRICES of Wq, Wk, Wv, Wo?

#5 opened 3 days ago by

这个版本对于5090单卡来说还是太大了

#4 opened 4 days ago by

Why is this 4bit version has a 32.7 GB size?

#3 opened 4 days ago by