Delete convd_host_weights and update all tests using conv2d #16264
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
shwetankTT
force-pushed
the
shwetankTT/conv_weight
branch
from
4b201bc to
cadb3dd
Compare
shwetankTT
changed the title
shwetankTT
changed the title
shwetankTT
requested review from
esmalTT,
uaydonat,
mywoodstock,
tt-aho,
dvartaniansTT,
sankarmanoj-tt and
pavlejosipovic
as code owners
| "batch_size": batch_size, | ||
| "input_height": input_tensor.shape[1], | ||
| "input_width": input_tensor.shape[2], | ||
| "kernel_size": (3, 3), |
There was a problem hiding this comment.
seems like most args are redundant between prepare_conv_weights and conv2d
I don't this we should hardcode params twice like kernel size (3, 3), all of these should be defined in one place.
It's weird that for prepare_conv_weights we use this kwargs dict, and on conv2d we use named argumets, I would align those.
Also, there is not need to define this conv_kwargs in case it's not going to be used, it should be under if where ttnn.prepare_conv_weights is.
This comment applies to all models changed.
| "conv_config": conv_config, | ||
| } | ||
|
|
||
| if True or not ttnn.is_tensor_storage_on_device(tt_weight): |
There was a problem hiding this comment.
Should "True or" part be removed?
| **conv_kwargs, | ||
| ) | ||
| self.bias = ( | ||
| ttnn.prepare_conv_bias( |
There was a problem hiding this comment.
prepare weights should be performed in init part of the model not in inference part
| } | ||
|
|
||
| if not ttnn.is_tensor_storage_on_device(self.weight): | ||
| breakpoint |
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Outdated suggestions cannot be applied.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.
Ticket
#14179
Problem description
Prepare conv2d_weight and bias in various models.
What's changed
Using Prepare_conv2d_weight and bias fuction in conv2d api.
Device perf regressions
Nightly model and ttnn tests
Model perf tests
Checklist