[jit] kill _parameter_list #27399
Closed
[jit] kill _parameter_list #27399
Conversation
This file contains hidden or 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
This was devised in a time when we didn't have module attributes. They are essentially just tensor lists, so represent them that way. This has the additional benefit of making the RNN forward pass faster because we effectively cache the flattened weights. The only complication part is that someone may come along and do: ``` my_rnn_mod.w_ih_l0 = torch.nn.Parameter(...) ``` This means we need to override setattr to keep the flattened weights cache up to date.
pytorchbot
added
oncall: jit
This was referenced Oct 4, 2019
Closed
Closed
zdevito
approved these changes
Oct 4, 2019
Contributor
zdevito
left a comment
zdevito
left a comment
There was a problem hiding this comment.
Thanks for cleaning this up.
torch/jit/quantized.py
Outdated
| if dtype == torch.int8: | ||
| self._orig_weights_names = self._packed_weights_names | ||
|
|
||
| self._packed_weights = torch.jit.Attribute([getattr(self, weight) for weight in self._packed_weights_names], List[Tensor]) |
Contributor
There was a problem hiding this comment.
Do we need to have torch.jit.Attribute or does it work to put the types in __attributes__?
Member
Author
There was a problem hiding this comment.
It needs to be an attribute with the current ScriptModule API, but not after my rewrite.
This was devised in a time when we didn't have module attributes. They are essentially just tensor lists, so represent them that way. This has the additional benefit of making the RNN forward pass faster because we effectively cache the flattened weights. The only complication part is that someone may come along and do: ``` my_rnn_mod.w_ih_l0 = torch.nn.Parameter(...) ``` This means we need to override setattr to keep the flattened weights cache up to date.
This was devised in a time when we didn't have module attributes. They are essentially just tensor lists, so represent them that way. This has the additional benefit of making the RNN forward pass faster because we effectively cache the flattened weights. The only complication part is that someone may come along and do: ``` my_rnn_mod.w_ih_l0 = torch.nn.Parameter(...) ``` This means we need to override setattr to keep the flattened weights cache up to date.
This was devised in a time when we didn't have module attributes. They are essentially just tensor lists, so represent them that way. This has the additional benefit of making the RNN forward pass faster because we effectively cache the flattened weights. The only complication part is that someone may come along and do: ``` my_rnn_mod.w_ih_l0 = torch.nn.Parameter(...) ``` This means we need to override setattr to keep the flattened weights cache up to date. Differential Revision: [D17785658](https://our.internmc.facebook.com/intern/diff/D17785658)
This was devised in a time when we didn't have module attributes. They are essentially just tensor lists, so represent them that way. This has the additional benefit of making the RNN forward pass faster because we effectively cache the flattened weights. The only complication part is that someone may come along and do: ``` my_rnn_mod.w_ih_l0 = torch.nn.Parameter(...) ``` This means we need to override setattr to keep the flattened weights cache up to date. Differential Revision: [D17785658](https://our.internmc.facebook.com/intern/diff/D17785658)
This was devised in a time when we didn't have module attributes. They are essentially just tensor lists, so represent them that way. This has the additional benefit of making the RNN forward pass faster because we effectively cache the flattened weights. The only complication part is that someone may come along and do: ``` my_rnn_mod.w_ih_l0 = torch.nn.Parameter(...) ``` This means we need to override setattr to keep the flattened weights cache up to date. Differential Revision: [D17785658](https://our.internmc.facebook.com/intern/diff/D17785658)
This was devised in a time when we didn't have module attributes. They are essentially just tensor lists, so represent them that way. This has the additional benefit of making the RNN forward pass faster because we effectively cache the flattened weights. The only complication part is that someone may come along and do: ``` my_rnn_mod.w_ih_l0 = torch.nn.Parameter(...) ``` This means we need to override setattr to keep the flattened weights cache up to date. Differential Revision: [D17785658](https://our.internmc.facebook.com/intern/diff/D17785658)
This was devised in a time when we didn't have module attributes. They are essentially just tensor lists, so represent them that way. This has the additional benefit of making the RNN forward pass faster because we effectively cache the flattened weights. The only complication part is that someone may come along and do: ``` my_rnn_mod.w_ih_l0 = torch.nn.Parameter(...) ``` This means we need to override setattr to keep the flattened weights cache up to date. Differential Revision: [D17785658](https://our.internmc.facebook.com/intern/diff/D17785658)
This was devised in a time when we didn't have module attributes. They are essentially just tensor lists, so represent them that way. This has the additional benefit of making the RNN forward pass faster because we effectively cache the flattened weights. The only complication part is that someone may come along and do: ``` my_rnn_mod.w_ih_l0 = torch.nn.Parameter(...) ``` This means we need to override setattr to keep the flattened weights cache up to date. Differential Revision: [D17785658](https://our.internmc.facebook.com/intern/diff/D17785658)
This was devised in a time when we didn't have module attributes. They are essentially just tensor lists, so represent them that way. This has the additional benefit of making the RNN forward pass faster because we effectively cache the flattened weights. The only complication part is that someone may come along and do: ``` my_rnn_mod.w_ih_l0 = torch.nn.Parameter(...) ``` This means we need to override setattr to keep the flattened weights cache up to date. Differential Revision: [D17785658](https://our.internmc.facebook.com/intern/diff/D17785658)
This was devised in a time when we didn't have module attributes. They are essentially just tensor lists, so represent them that way. This has the additional benefit of making the RNN forward pass faster because we effectively cache the flattened weights. The only complication part is that someone may come along and do: ``` my_rnn_mod.w_ih_l0 = torch.nn.Parameter(...) ``` This means we need to override setattr to keep the flattened weights cache up to date. Differential Revision: [D17785658](https://our.internmc.facebook.com/intern/diff/D17785658)
Contributor
1 similar comment
Contributor
thiagocrepaldi
pushed a commit
to thiagocrepaldi/pytorch
that referenced
this pull request
Feb 4, 2020
Summary: Pull Request resolved: pytorch#27399 This was devised in a time when we didn't have module attributes. They are essentially just tensor lists, so represent them that way. This has the additional benefit of making the RNN forward pass faster because we effectively cache the flattened weights. The only complication part is that someone may come along and do: ``` my_rnn_mod.w_ih_l0 = torch.nn.Parameter(...) ``` This means we need to override setattr to keep the flattened weights cache up to date. Test Plan: Imported from OSS Differential Revision: D17785658 Pulled By: suo fbshipit-source-id: 7789cd1d0d4922bfd5eba1716976442fbf150766
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.
Stack from ghstack:
This was devised in a time when we didn't have module attributes. They
are essentially just tensor lists, so represent them that way. This has
the additional benefit of making the RNN forward pass faster because we
effectively cache the flattened weights.
The only complication part is that someone may come along and do:
This means we need to override setattr to keep the flattened weights
cache up to date.
Differential Revision: D17785658