Add qk norm optionally before attention calculation #8820
Merged
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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/8820
Note: Links to docs will display an error until the docs builds have been completed. ❌ 2 Cancelled JobsAs of commit 5b2587a with merge base 73acde9 ( CANCELLED JOBS - The following jobs were cancelled. Please retry:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D70355802 |
|
@pytorchbot label "topic: not user facing" |
madhu-fb
force-pushed
the
export-D70355802
branch
from
e932c2d to
70f8a2d
Compare
madhu-fb
added a commit
to madhu-fb/executorch
that referenced
this pull request
Mar 5, 2025
Summary: Some of the new llama checkpoints developed by genai use an additional qk_norm in the attention calculation. To run these models with executorch and have parity with server models, we require an optional qk norm in the ET attention. Refactoring RMSNorm into a separate file so that there is no circular dependency between attention and llama_transformer Differential Revision: D70355802
|
This pull request was exported from Phabricator. Differential Revision: D70355802 |
madhu-fb
added a commit
to madhu-fb/executorch
that referenced
this pull request
Mar 5, 2025
Summary: Pull Request resolved: pytorch#8820 Some of the new llama checkpoints developed by genai use an additional qk_norm in the attention calculation. To run these models with executorch and have parity with server models, we require an optional qk norm in the ET attention. Refactoring RMSNorm into a separate file so that there is no circular dependency between attention and llama_transformer Differential Revision: D70355802
madhu-fb
force-pushed
the
export-D70355802
branch
from
70f8a2d to
4d64a51
Compare
madhu-fb
added a commit
to madhu-fb/executorch
that referenced
this pull request
Mar 6, 2025
Summary: Some of the new llama checkpoints developed by genai use an additional qk_norm in the attention calculation. To run these models with executorch and have parity with server models, we require an optional qk norm in the ET attention. Refactoring RMSNorm into a separate file so that there is no circular dependency between attention and llama_transformer Reviewed By: iseeyuan Differential Revision: D70355802
madhu-fb
force-pushed
the
export-D70355802
branch
from
4d64a51 to
44f9842
Compare
madhu-fb
added a commit
to madhu-fb/executorch
that referenced
this pull request
Mar 6, 2025
Summary: Some of the new llama checkpoints developed by genai use an additional qk_norm in the attention calculation. To run these models with executorch and have parity with server models, we require an optional qk norm in the ET attention. Refactoring RMSNorm into a separate file so that there is no circular dependency between attention and llama_transformer Reviewed By: iseeyuan Differential Revision: D70355802
madhu-fb
force-pushed
the
export-D70355802
branch
from
44f9842 to
f850b04
Compare
|
This pull request was exported from Phabricator. Differential Revision: D70355802 |
madhu-fb
added a commit
to madhu-fb/executorch
that referenced
this pull request
Mar 6, 2025
Summary: Pull Request resolved: pytorch#8820 Some of the new llama checkpoints developed by genai use an additional qk_norm in the attention calculation. To run these models with executorch and have parity with server models, we require an optional qk norm in the ET attention. Refactoring RMSNorm into a separate file so that there is no circular dependency between attention and llama_transformer Reviewed By: iseeyuan Differential Revision: D70355802
madhu-fb
force-pushed
the
export-D70355802
branch
from
f850b04 to
66f92c5
Compare
madhu-fb
added a commit
to madhu-fb/executorch
that referenced
this pull request
Mar 6, 2025
Summary: Pull Request resolved: pytorch#8820 Some of the new llama checkpoints developed by genai use an additional qk_norm in the attention calculation. To run these models with executorch and have parity with server models, we require an optional qk norm in the ET attention. Refactoring RMSNorm into a separate file so that there is no circular dependency between attention and llama_transformer Reviewed By: iseeyuan Differential Revision: D70355802
madhu-fb
force-pushed
the
export-D70355802
branch
from
66f92c5 to
90b24b1
Compare
|
This pull request was exported from Phabricator. Differential Revision: D70355802 |
madhu-fb
force-pushed
the
export-D70355802
branch
from
90b24b1 to
0cb1641
Compare
madhu-fb
added a commit
to madhu-fb/executorch
that referenced
this pull request
Mar 6, 2025
Summary: Some of the new llama checkpoints developed by genai use an additional qk_norm in the attention calculation. To run these models with executorch and have parity with server models, we require an optional qk norm in the ET attention. Refactoring RMSNorm into a separate file so that there is no circular dependency between attention and llama_transformer Reviewed By: iseeyuan Differential Revision: D70355802
|
This pull request was exported from Phabricator. Differential Revision: D70355802 |
madhu-fb
added a commit
to madhu-fb/executorch
that referenced
this pull request
Mar 6, 2025
Summary: Pull Request resolved: pytorch#8820 Some of the new llama checkpoints developed by genai use an additional qk_norm in the attention calculation. To run these models with executorch and have parity with server models, we require an optional qk norm in the ET attention. Refactoring RMSNorm into a separate file so that there is no circular dependency between attention and llama_transformer Reviewed By: iseeyuan Differential Revision: D70355802
madhu-fb
force-pushed
the
export-D70355802
branch
2 times, most recently
from
51a970b to
388f54d
Compare
madhu-fb
added a commit
to madhu-fb/executorch
that referenced
this pull request
Mar 6, 2025
Summary: Some of the new llama checkpoints developed by genai use an additional qk_norm in the attention calculation. To run these models with executorch and have parity with server models, we require an optional qk norm in the ET attention. Refactoring RMSNorm into a separate file so that there is no circular dependency between attention and llama_transformer Reviewed By: iseeyuan Differential Revision: D70355802
Summary: Pull Request resolved: pytorch#8820 Some of the new llama checkpoints developed by genai use an additional qk_norm in the attention calculation. To run these models with executorch and have parity with server models, we require an optional qk norm in the ET attention. Refactoring RMSNorm into a separate file so that there is no circular dependency between attention and llama_transformer Reviewed By: iseeyuan Differential Revision: D70355802
|
This pull request was exported from Phabricator. Differential Revision: D70355802 |
madhu-fb
force-pushed
the
export-D70355802
branch
from
388f54d to
5b2587a
Compare
# for free
to join this conversation on GitHub.
Already have an account?
# 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.
Differential Revision: D70355802