dongyangli-del / EEG_Image_decode

Using vision-language models to decode natural image perception from non-invasive brain recordings.
MIT License
78 stars 15 forks source link

huggingface Network address error #15

Closed koro-github closed 3 months ago

koro-github commented 3 months ago

when I was running ATM_S_reconstruction_scale_0_1000.py, an error occurred in the eegdatasets_leaveone.py file. I checked the cause online and said that the address of huggingface was incorrect. c11fc7b85a902e7d518ee40660677f2

dongyangli-del commented 3 months ago

Hi, @koro-github , If you are in mainland China, you need to set up a proxy based on your VPN. This means you need to fill in an IP address and port number with a foreign proxy:

proxy ='http://xx.xx.xx.xx:port'
os.environ['http_proxy']= proxy
os.environ['https_proxy']= proxy

But if you are in other regions outside of mainland China, these lines of code are unnecessary.

koro-github commented 3 months ago

Hi Dongyang, I have successfully run ATM_S_reconstruction_scale_0_1000.py and saved the weights. Now I am trying to run your Generation_metrics_sub1.ipynb, but there are some problems, some packages are not imported, can you tell me how to import them? 94652617a685d929267d5556b19df6d 41786390a395be461d09c5240aeb846

dongyangli-del commented 3 months ago

These classes or functions are defined in the custom_pipeline.py module. And if you need them, you should uncomment this code and import these modules: image

koro-github commented 3 months ago

Thanks for your answer. Now the problem of import package has been solved, but I still can't find the list of texts. 1.Can you tell me how to find or define a list of texts? 2.Generation_metrics_sub1.ipynb file contains train prior diffusion. Could you please tell me the correct reading order of ipynb file? 3f22af9766c267da96e893a95aca20d

dongyangli-del commented 3 months ago
  1. I assume that the list of texts you mentioned refers to the text description of each image. Here, the image name label is simply used.
  2. Generation_metrics_sub1.ipynb is missing the definition of texts, I have updated it. Other files are not missing, you can try using other files first.
koro-github commented 3 months ago

thanks you answer,dongyang. I can't find the difference between ATM_S_insubject_retrieval.py and ATM_S_insubject_classification.py. One of them is retrieval and the other is classification. What's the difference in the code. there model 、acc_compute、dataset are same. ATM_S_insubject_classification.py 4dffb4e6a25e332b44f10d189d2e44a

ATM_S_insubject_classification.py a26f4edf40dfdc87db5f2cbe9e79657