Thank you for sharing your code! I'm reproducing BotRGCN on dataset Twibot-20using your code. However, after running "preprocess_1.py", "the line cat_properties_tensor=torch.reshape(default_profile_image_tensor,(5301,1)) " has error "RuntimeError: shape '[5301, 1]' is invalid for input of size 229580 ".
After running "preprocess_2.py" the line " each_tweet_tensor=torch.tensor(feature_extract(each_tweet)) " has error “ RuntimeError: The expanded size of the tensor (710) must match the existing size (514) at non-singleton dimension 1. Target sizes: [1, 710]. Tensor sizes: [1, 514]”
After running "train.py" the line “output = model(des_tensor,tweets_tensor,num_prop,category_prop,edge_index,edge_type)” has error
RuntimeError: mat1 and mat2 shapes cannot be multiplied (229580x1 and 3x8)
I have tried many modifications without success, could you please answer my doubts and thank you again!
Thank you for sharing your code! I'm reproducing BotRGCN on dataset Twibot-20using your code. However, after running "preprocess_1.py", "the line cat_properties_tensor=torch.reshape(default_profile_image_tensor,(5301,1)) " has error "RuntimeError: shape '[5301, 1]' is invalid for input of size 229580 ".
After running "preprocess_2.py" the line " each_tweet_tensor=torch.tensor(feature_extract(each_tweet)) " has error “ RuntimeError: The expanded size of the tensor (710) must match the existing size (514) at non-singleton dimension 1. Target sizes: [1, 710]. Tensor sizes: [1, 514]”
After running "train.py" the line “output = model(des_tensor,tweets_tensor,num_prop,category_prop,edge_index,edge_type)” has error RuntimeError: mat1 and mat2 shapes cannot be multiplied (229580x1 and 3x8)
I have tried many modifications without success, could you please answer my doubts and thank you again!