Refactor MLP #3257
Merged
Refactor MLP #3257
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
arunppsg
approved these changes
Feb 28, 2023
rbharath
reviewed
Mar 5, 2023
| self.model = nn.Sequential(*self.build_layers()) | ||
| self.skip = nn.Linear(d_input, d_output) if skip_connection else None | ||
|
|
||
| def build_layers(self): |
| input = x | ||
| for layer in self.model: | ||
| x = layer(x) | ||
| if isinstance(layer, nn.Linear): |
There was a problem hiding this comment.
Maybe better to add the activations into the sequential model directly? Is this style off adding activations via a loop standard elsewhere?
There was a problem hiding this comment.
If we do that, we will have to use the torch activations instead of the deepchem activations, since you cannot make a sequential model with anything but nn.modules.
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Outdated suggestions cannot be applied.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.
Pull Request Template
Description
This PR allows the user to create an MLP with different sized hidden layers using a tuple, which was not previously possible. It also adds an activation on the final layer which would likely be expected. This change was made in an effort to unify the different feedforward network implementations in deepchem.
This PR also fixes a bug in the case of selecting to use 1 hidden layer. The current implementation produces a layer in the shape of [input_dim,hidden_dim], not the expected [input_dim,output_dim].
Type of change
Please check the option that is related to your PR.
Checklist
yapf -i <modified file>and check no errors (yapf version must be 0.32.0)mypy -p deepchemand check no errorsflake8 <modified file> --countand check no errorspython -m doctest <modified file>and check no errors