Timeline for Classifier using pytorch
Current License: CC BY-SA 4.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 21, 2020 at 1:01 | vote | accept | Blade | ||
| Jun 21, 2020 at 1:01 | comment | added | Blade | Nvm, I just set the dimensions wrong. Thank you! | |
| Jun 21, 2020 at 0:34 | comment | added | Blade |
It seems that in order for this predictor to work, I would have to know the batch size in advance. In other words, when doing torch.matmul(self.weights, features) - self.threshold I need the batch size to set the dimensionality of self.weights in advance. However, when using nn.Linear(10, 1) I don't have to deal with the batch size anymore. Is this true or should I change something?
|
|
| Jun 20, 2020 at 21:10 | comment | added | SrJ | Refer to this article. You will find detailed explanation. stackoverflow.com/a/51027227/11961138 | |
| Jun 20, 2020 at 21:06 | comment | added | Blade |
Thanks, could you alter my class predictor(nn.Module) with nn.Parameter() so I can understand?
|
|
| Jun 20, 2020 at 20:51 | history | answered | SrJ | CC BY-SA 4.0 |