deepseek-ai

/

Janus-Pro-7B

like 2.1k

Follow

DeepSeek 21.1k

parallel_size: int = 16
Changing this to 1 used ~16gb max

In the generation inference.py file?

Yeah :3

Line 60 i believe

Yeah, but for me, while loading only it's taking more than 30 GB.

Hi, can you please share how you setup the model? I can't seem to use the model using transformer, it complains about the not knowing the model

Hi, can you please share how you setup the model? I can't seem to use the model using transformer, it complains about the not knowing the model

You can find instructions here: https://github.com/deepseek-ai/Janus?tab=readme-ov-file#janus-pro
They worked for me with no issues

Hi, can you please share how you setup the model? I can't seem to use the model using transformer, it complains about the not knowing the model

You can find instructions here: https://github.com/deepseek-ai/Janus?tab=readme-ov-file#janus-pro
They worked for me with no issues

thanks!