nvidia
/

Gemma-4-31B-IT-NVFP4

Text Generation

Model Optimizer

Model card Files Files and versions

Resources

View closed (7)

Update tokenizer_config.json to adhere to the upstream version of google/gemma-4-31B-it

#18 opened 6 days ago by

fix(chat_template): backport latest Google canonical template

#16 opened about 1 month ago by

taniguchimasaya

Model tree shows "Finetuned" instead of "Quantized"

#15 opened about 2 months ago by

How to run this model YAML with sparkrun on NVIDIA GB10

#14 opened 2 months ago by

This model wasn't trained with FP4 or NVFP4

#8 opened 5 months ago by

1*H100 with vLLM 0.19.0 Failed

#7 opened 5 months ago by

Question about q_scale / KV cache scale fallback in vLLM for Gemma-4-31B-IT-NVFP4: expected accuracy impact?

#6 opened 5 months ago by

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

#5 opened 5 months ago by

这个版本对于5090单卡来说还是太大了

#4 opened 5 months ago by

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

#3 opened 5 months ago by