[Concept Entry] PyTorch: Built-in Loss Functions #6085
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
avdhoottt
added
new entry
avdhoottt
requested changes
Feb 4, 2025
There was a problem hiding this comment.
Hey @PragatiVerma18, I've reviewed the entry and left some comments. Please make the changes, thank you!!
Comment on lines
+61
to
+63
| - `size_average` (bool, optional): Deprecated (use `reduction` instead). If `True`, the loss is averaged over observations for each mini-batch. Default is `True`. | ||
| - `ignore_index` (int, optional): Specifies a target value that is ignored and does not contribute to the loss calculation. | ||
| - `reduce` (bool, optional): Deprecated (use `reduction` instead). If `True`, it sums the losses across the batch. Default is `True`. |
There was a problem hiding this comment.
If these are deprecated, then I think we don't need to include them.
| predictions = torch.tensor([2.0, 3.0, 4.0]) | ||
| targets = torch.tensor([2.5, 3.5, 4.5]) | ||
| loss = loss_fn(predictions, targets) | ||
| print(loss) # Output: tensor(0.0833) |
There was a problem hiding this comment.
Recheck the output of this code here. I'm getting this output - tensor(0.2500)
| logits = torch.tensor([[1.0, 2.0, 3.0], [1.0, 2.0, 3.0]]) | ||
| labels = torch.tensor([2, 0]) | ||
| loss = loss_fn(logits, labels) | ||
| print(loss) # Output: tensor(0.4076) |
There was a problem hiding this comment.
I'm getting this output for the above code - tensor(1.4076)
| logits = torch.tensor([0.5, -1.5, 2.0]) | ||
| labels = torch.tensor([1.0, 0.0, 1.0]) | ||
| loss = loss_fn(logits, labels) | ||
| print(loss) # Output: tensor(0.4891) |
Comment on lines
+166
to
+175
| import torch | ||
| import torch.nn as nn | ||
|
|
||
| # Example of CosineEmbeddingLoss | ||
| loss_fn = nn.CosineEmbeddingLoss() | ||
| input1 = torch.tensor([1.0, 0.0]) | ||
| input2 = torch.tensor([0.0, 1.0]) | ||
| target = torch.tensor([1]) # 1 means the inputs should be similar | ||
| loss = loss_fn(input1, input2, target) | ||
| print(loss) # Output: tensor(2.0) |
There was a problem hiding this comment.
It's giving a runtime error. Please correct the code.
Comment on lines
+198
to
+207
| import torch | ||
| import torch.nn as nn | ||
|
|
||
| # Example of KLDivLoss | ||
| loss_fn = nn.KLDivLoss(reduction='batchmean') | ||
| input = torch.tensor([[0.0, 0.1, 0.2], [0.1, 0.0, 0.3]]) | ||
| target = torch.tensor([[0.0, 0.1, 0.3], [0.1, 0.2, 0.2]]) | ||
| loss = loss_fn(input.log(), target) | ||
| print(loss) # Output: tensor(0.0237) | ||
| ``` |
There was a problem hiding this comment.
Something wrong with the code snippet. It's giving tensor(nan) output.
|
|
||
| ## Choosing the Right Loss Function | ||
|
|
||
| When selecting a loss function for your model, it is essential to consider the task you're working on. |
There was a problem hiding this comment.
Please don't use first and second-person pronouns.
avdhoottt
added
status: review 1️⃣ completed
status: waiting for author
and removed
status: under review
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Add a new entry on Built-in Loss Functions concept under PyTorch.
Issue Solved
Closes #5861
Type of Change
Checklist
mainbranch.Issues Solvedsection.