search

SiyuanHuang95 / ManipVQA

ManipVQA: Injecting Robotic Affordance and Physically Grounded Information into Multi-Modal Large Language Models
40 stars 1 forks source link

image_test #1

Closed GentlesJan closed 1 month ago

GentlesJan commented 2 months ago

Hi!

How can I quickly test the performance of your pre-trained model on new images?

Thanks!

SiyuanHuang95 commented 2 months ago

Hi, thanks for your interest in our project.

To run our model, you need 2-A100 GPUs if you want to run locally.

If you just wanna give a quick trial (limited images), you can send the images and the corresponding question prompts in a text file, and I would try to run it on the lab's cluster. You can send me via email: anthonyhuang4de@gmail.com

SiyuanHuang95 commented 1 month ago

Since no following message, I will close this issue. You can reopen it whenever you need it. @GentlesJan