Cast long inputs to float32 before feeding to xnnpack graph runtime #553
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
✅ Deploy Preview for resplendent-gnome-14e531 ready!
To edit notification comments on pull requests, go to your Netlify site configuration. |
facebook-github-bot
added
the
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D49797819 |
mcr229
force-pushed
the
export-D49797819
branch
from
8deb0b2 to
53bd3c0
Compare
|
This pull request was exported from Phabricator. Differential Revision: D49797819 |
mcr229
force-pushed
the
export-D49797819
branch
from
53bd3c0 to
6ad27d5
Compare
|
This pull request was exported from Phabricator. Differential Revision: D49797819 |
…#553) Summary: In DeepLab v3, we saw that some inputs are int64, (this is largely from interpolation decomposition). XNNPACK can not handle int64, which means that the int64 inputs would wrongly be interpreted as float32. In order to correctly feed the right inputs to XNNPACK, we must cast the int64 data to float32. We do something similar when casting int32 data to int64 when we are returning indicies from argmax pooling. Reviewed By: digantdesai Differential Revision: D49797819
mcr229
force-pushed
the
export-D49797819
branch
from
6ad27d5 to
72f6766
Compare
|
This pull request was exported from Phabricator. Differential Revision: D49797819 |
|
This pull request has been merged in 7906c18. |
tarun292
pushed a commit
that referenced
this pull request
Oct 5, 2023
…553) Summary: Pull Request resolved: #553 In DeepLab v3, we saw that some inputs are int64, (this is largely from interpolation decomposition). XNNPACK can not handle int64, which means that the int64 inputs would wrongly be interpreted as float32. In order to correctly feed the right inputs to XNNPACK, we must cast the int64 data to float32. We do something similar when casting int32 data to int64 when we are returning indicies from argmax pooling. Reviewed By: digantdesai Differential Revision: D49797819 fbshipit-source-id: 1f4f6ab38c4c8e36ba0a7017300f5e3a3ccbf810
Gasoonjia
pushed a commit
that referenced
this pull request
Jul 30, 2024
* test sdpa with fp16 * kv cache fp32 * typo
Gasoonjia
pushed a commit
that referenced
this pull request
Jul 30, 2024
* make --device fast the default * Update iOS.md (#517) * Update iOS.md * Update iOS.md * Pip to pip3 (#504) * remove macos-12 test * pip to pip3 * break aoti CI jobs separately (#500) * init * fixes * more fixes * fixes * fix * fix * bug fix * add objcopy update * suppress int8 * undefined variable --------- Co-authored-by: Michael Gschwind <[email protected]> * Support llama3 in chat in run.cpp (#486) * refactor chat runner in preparation for llama3 * add sketch for llama3 prompt template and move to returning tokens * fix tiktoken * fixes to chat * add default llama_ver * Add tests for quantize json, add cuda device specification and precision to cuda.json (#519) * remove code for no KV Cache path (#527) * Update ADVANCED-USERS.md (#529) Update Advanced Users description to reflect changes in the repo since the description was initially created. * runner-aoti on cuda (#531) * runner-aoti on cuda * transfer results back to CPU * transfer results back to CPU * runner-aoti on cuda * Update runner_build.md (#530) Update description of runner and build process in runner_build.md * clean up runner code a little (#532) * clean up runner code a little * update * update * pull out generate loop in chat * updates * edit docs * typo * move int8 linear class and function into qops.py (#534) * add dtype tests for runner-aoti + runner-et (#539) * add dtype tests for runner-aoti + runner-et * typo * Quantized embedding (#536) * move int8 linear class and function into qops.py * move Quantized Embedding to qops.py * Move Linear int4 to qops (#537) * move int8 linear class and function into qops.py * move Quantized Embedding to qops.py * move int4 linear to qops * Revert "add dtype tests for runner-aoti + runner-et (#539)" (#548) This reverts commit a7a24577a65be67ac9ae4dc05452f35d9c49e5d1. * fix generate for llama3 (#538) * fix generate for llama3 * switch more things to C * remove C++ header * add delegation visualization instructions (#551) * Add dtype runner aoti (#552) * add dtype tests for runner-aoti + runner-et * typo * add dtype test runner-aoti * test sdpa with fp16 (#553) * test sdpa with fp16 * kv cache fp32 * typo * update (#560) * Only support newest versions of lm-eval (#556) Summary: remove support for lm-eval 0.3 to reduce the options we have Test Plan: CI Reviewers: Subscribers: Tasks: Tags: * split cpu eval CI by dtype (#554) * split cpu eval CI by dtype * fix * differentiate names with checks * keep one name the same as old * fix * Removing duplicate HF issue message from README (#559) Co-authored-by: Michael Gschwind <[email protected]> * doc updates (#567) * Add VM-safe MPS check --------- Co-authored-by: Anthony Shoumikhin <[email protected]> Co-authored-by: metascroy <[email protected]> Co-authored-by: Nikita Shulga <[email protected]> Co-authored-by: lucylq <[email protected]> Co-authored-by: Jerry Zhang <[email protected]> Co-authored-by: Jack-Khuu <[email protected]>
Gasoonjia
pushed a commit
that referenced
this pull request
Jul 30, 2024
* code beautification * code beautification, move functions together * make --device fast the default (#515) * make --device fast the default * Update iOS.md (#517) * Update iOS.md * Update iOS.md * Pip to pip3 (#504) * remove macos-12 test * pip to pip3 * break aoti CI jobs separately (#500) * init * fixes * more fixes * fixes * fix * fix * bug fix * add objcopy update * suppress int8 * undefined variable --------- Co-authored-by: Michael Gschwind <[email protected]> * Support llama3 in chat in run.cpp (#486) * refactor chat runner in preparation for llama3 * add sketch for llama3 prompt template and move to returning tokens * fix tiktoken * fixes to chat * add default llama_ver * Add tests for quantize json, add cuda device specification and precision to cuda.json (#519) * remove code for no KV Cache path (#527) * Update ADVANCED-USERS.md (#529) Update Advanced Users description to reflect changes in the repo since the description was initially created. * runner-aoti on cuda (#531) * runner-aoti on cuda * transfer results back to CPU * transfer results back to CPU * runner-aoti on cuda * Update runner_build.md (#530) Update description of runner and build process in runner_build.md * clean up runner code a little (#532) * clean up runner code a little * update * update * pull out generate loop in chat * updates * edit docs * typo * move int8 linear class and function into qops.py (#534) * add dtype tests for runner-aoti + runner-et (#539) * add dtype tests for runner-aoti + runner-et * typo * Quantized embedding (#536) * move int8 linear class and function into qops.py * move Quantized Embedding to qops.py * Move Linear int4 to qops (#537) * move int8 linear class and function into qops.py * move Quantized Embedding to qops.py * move int4 linear to qops * Revert "add dtype tests for runner-aoti + runner-et (#539)" (#548) This reverts commit a7a24577a65be67ac9ae4dc05452f35d9c49e5d1. * fix generate for llama3 (#538) * fix generate for llama3 * switch more things to C * remove C++ header * add delegation visualization instructions (#551) * Add dtype runner aoti (#552) * add dtype tests for runner-aoti + runner-et * typo * add dtype test runner-aoti * test sdpa with fp16 (#553) * test sdpa with fp16 * kv cache fp32 * typo * update (#560) * Only support newest versions of lm-eval (#556) Summary: remove support for lm-eval 0.3 to reduce the options we have Test Plan: CI Reviewers: Subscribers: Tasks: Tags: * split cpu eval CI by dtype (#554) * split cpu eval CI by dtype * fix * differentiate names with checks * keep one name the same as old * fix * Removing duplicate HF issue message from README (#559) Co-authored-by: Michael Gschwind <[email protected]> * doc updates (#567) * Add VM-safe MPS check --------- Co-authored-by: Anthony Shoumikhin <[email protected]> Co-authored-by: metascroy <[email protected]> Co-authored-by: Nikita Shulga <[email protected]> Co-authored-by: lucylq <[email protected]> Co-authored-by: Jerry Zhang <[email protected]> Co-authored-by: Jack-Khuu <[email protected]> * add unpacking support (#525) * add unpacking support * fix typos and linter * perform parallel prefill when possible (#568) * perform parallel prefill when possible * typo * disable hack * remove print * remove debug messages which prevent export * fixes * stream results in generate.py (#571) * remove logging interfering with export --------- Co-authored-by: Anthony Shoumikhin <[email protected]> Co-authored-by: metascroy <[email protected]> Co-authored-by: Nikita Shulga <[email protected]> Co-authored-by: lucylq <[email protected]> Co-authored-by: Jerry Zhang <[email protected]> Co-authored-by: Jack-Khuu <[email protected]>
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.
Summary:
In DeepLab v3, we saw that some inputs are int64, (this is largely from interpolation decomposition). XNNPACK can not handle int64, which means that the int64 inputs would wrongly be interpreted as float32.
In order to correctly feed the right inputs to XNNPACK, we must cast the int64 data to float32. We do something similar when casting int32 data to int64 when we are returning indicies from argmax pooling.
Differential Revision: D49797819