Shashin
The label for reposnse in train_dataloader should be Long. In `train.py` ```python labels = labels.to(dtype=torch.long).to(device) ```
```python self.classifier = nn.Sequential( nn.Flatten(), nn.Linear(in_features=hidden_units, out_features=output_shape ) ``` The linear _input_feat_ (matrix 2 no of rows) should match your 2nd conv2d _output_feat_ (matrix 1 no of column) --- ```...
```bash pip3 install torch ```
The problem here is its just sample data here. More api calls need to be made for actual data and there are around 35000 (rough estimate) schools in Nepal. Don't...
For downloading feature I don't think columns like longitude and latitude are needed👍🏻 Also add note specifying what is scode and cscode at the bottom in pdf
> * Allow the user to view the downloaded directly from the UI itself. A simple toast with file URL should do the trick. I'm not sure though. By this...
Don't think this is necessary since it is basic command to install pip packages.
I'm not sure if mobile applications are within the scope, but once we have minimized bugs in the main code and streamlined the Streamlit web applications, we can consider developing...
Try doing from terminal instead ```python python >> import nltk >> nltk.download('punkt') ``` If it doesn't throw error Coment the `nltk.download('punkt')` part
Same error here as well