problem in dataloader.py

[DimPenCHEN]

I have solved the data problems(I think),but I found that the file "dataloader.py" using the "label" key in data.json which is no in my data.json
self.data_complete = self.data_complete[self.data_complete['label']!=2] # label: 0-real, 1-fake, 2-debunk

BTW, I haven't obatined the data directory "vids" and **"ptvgg19_frame_thumb"**which aren't mentioned in the public dataset.
I wonder how to sovle this problem in dataloader.py TwT

[pp-jia]

How did you obtain the Dataset? Did you crawl it based on data.json? @DimPenCHEN

[DimPenCHEN]

How did you obtain the Dataset? Did you crawl it based on data.json? @DimPenCHEN

I obtaioned the data from the public dataset which in huggingface.co. I can send the detail address to you later(I can't find it right now).
Maybe We can exchange the contact information，we can communicate with this project more.

[DimPenCHEN]

@pp-jia https://huggingface.co/datasets/MischaQI/FakeSV/tree/main
This link is available to obtain the data you need

[pp-jia]

Thanks, you can contact me via the email address on my homepage. @DimPenCHEN

[TODO-main]

I have solved the data problems(I think),but I found that the file "dataloader.py" using the "label" key in data.json which is no in my data.json self.data_complete = self.data_complete[self.data_complete['label']!=2] # label: 0-real, 1-fake, 2-debunk

BTW, I haven't obatined the data directory "vids" and **"ptvgg19_frame_thumb"**which aren't mentioned in the public dataset. I wonder how to sovle this problem in dataloader.py TwT

I have the same problem, have you solved this problem

[TODO-main]

I have solved the data problems(I think),but I found that the file "dataloader.py" using the "label" key in data.json which is no in my data.json self.data_complete = self.data_complete[self.data_complete['label']!=2] # label: 0-real, 1-fake, 2-debunk
BTW, I haven't obatined the data directory "vids" and **"ptvgg19_frame_thumb"**which aren't mentioned in the public dataset. I wonder how to sovle this problem in dataloader.py TwT

I have the same problem, have you solved this problem

There is no ”lable“ attribute in the supplied "data.json" file

[andr2w]

The data.json file does not contain the 'label' column.
My way to solve this problem is to modify the SVFENDDataset as follows:

replace_values = {'辟谣': 2, '假': 1, '真':0}
self.data_complete['annotation'] = self.data_complete['annotation'].replace(replace_values)
self.data_complete = self.data_complete[self.data_complete['annotation']!=2]
 #self.data_complete = self.data_complete[self.data_complete['label']!=2] # label: 0-real, 1-fake, 2-debunk

[andr2w]

The data.json file does not contain the 'label' column. My way to solve this problem is to modify the SVFENDDataset as follows:

replace_values = {'辟谣': 2, '假': 1, '真':0}
self.data_complete['annotation'] = self.data_complete['annotation'].replace(replace_values)
self.data_complete = self.data_complete[self.data_complete['annotation']!=2]
 #self.data_complete = self.data_complete[self.data_complete['label']!=2] # label: 0-real, 1-fake, 2-debunk

If you follow the above, be sure to modify the following code into

label = 0 if item['annotation'] == '真' else 1

to

label = 0 if item['annotation'] == 0 else  1

Otherwise, there will be a significant error during the training phase.

[TODO-main]

The data.json file does not contain the 'label' column. My way to solve this problem is to modify the SVFENDDataset as follows:

replace_values = {'辟谣': 2, '假': 1, '真':0}
self.data_complete['annotation'] = self.data_complete['annotation'].replace(replace_values)
self.data_complete = self.data_complete[self.data_complete['annotation']!=2]
 #self.data_complete = self.data_complete[self.data_complete['label']!=2] # label: 0-real, 1-fake, 2-debunk

If you follow the above, be sure to modify the following code into

label = 0 if item['annotation'] == '真' else 1

to

label = 0 if item['annotation'] == 0 else  1

Otherwise, there will be a significant error during the training phase.

Thank you for your reply, which gave me a good solution to this problem. However, after solving this problem, I met a new problem. I wonder if you have also met the same problem?