Small l opt c++ #4508

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Open

kudomcho wants to merge 25 commits into pytorch:main from kudomcho:small-l-opt-c++

+723 −333

kudomcho commented Jul 16, 2025

No description provided.

kudomcho added 18 commits

January 30, 2025 12:17


          fixed find_max_Ls function to return int type

91a4559


          removed prints on max_ls vars

139b529


          Merge branch 'pytorch:main' into main

8fdb98b


          Merge branch 'pytorch:main' into main

d23685c


          optimized inference for L=1 case

cdf3977


          fixed packed_bags logic on max_L and added new condition on find_max_ls

f3654d6


          adapted linter on None condition

43114ad


          added args of max_ls on codegen_lookup_func

30095f7


          fixed errors on flake8

b93ebac


          fixed the relative path on common on test scripts

c889646


          formatted ufmt and fixed flake8 on nbit_split_embeddings_test.py


          formatted ufmt and fixed flake8 on nbit_split_embeddings_test.py

65c77da


          changed datatype to optional int64 with default values

8e9b7d9


          removed Ls arg for benchmark scripts

a26a48f


          supported nan weight handling

59708f9


          removed Ls args

971e8bc


          removed Ls args on test script

fb0d6b3


          fixed nan weight test and ',' on host.cpp

600df74

netlify bot commented Jul 16, 2025 •

edited

Loading

✅ Deploy Preview for pytorch-fbgemm-docs ready!

Name	Link
🔨 Latest commit	`a599f18`
🔍 Latest deploy log	https://app.netlify.com/projects/pytorch-fbgemm-docs/deploys/687e850f9b009d0008ee3232
😎 Deploy Preview	https://deploy-preview-4508--pytorch-fbgemm-docs.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

To edit notification comments on pull requests, go to your Netlify project configuration.

meta-cla bot added the cla signed label

kudomcho added 3 commits

July 16, 2025 21:28


          updated cpu_template and adjusted args position on host_cpu

5f6cbd2


          fixed 1 space on all lines

49d9373


          Merge branch 'main' into small-l-opt-c++

70ae79a

q10 reviewed

View reviewed changes

fbgemm_gpu/codegen/inference/embedding_forward_quantized_split_nbit_host_template.cu

                   #ifdef X
                   #undef X
                   #endif
-                  #define X(DeviceOnly, PackedMode, OutputRowsPerThread, InputRowsInFlight, MinNum128BRows, MaxNum128BRows) \
-                    FBGEMM_LAUNCH_KERNEL( \

Contributor

q10 Jul 17, 2025 •

edited

Loading

@kudomcho Please continue to use the FBGEMM_LAUNCH_KERNEL and PTA_B macros, which provide a bunch of runtime checks.

Author

kudomcho Jul 17, 2025

@q10 Addressed and keep the .BGEMM_LAUNCH_KERNEL and PTA_B macros,

Contributor

facebook-github-bot commented Jul 17, 2025

@q10 has imported this pull request. If you are a Meta employee, you can view this in D78466075.

kudomcho added 2 commits

July 17, 2025 16:41


          added F401 on nvfp4 import on quantize_ops

94aaf4e


          remain the fbgemm_kernel_launch

86ad794

Contributor

facebook-github-bot commented Jul 17, 2025

@q10 has imported this pull request. If you are a Meta employee, you can view this in D78466075.


           remove unused import

dda9e4d

Contributor

facebook-github-bot commented Jul 17, 2025

@q10 has imported this pull request. If you are a Meta employee, you can view this in D78466075.

Contributor

facebook-github-bot commented Jul 18, 2025

@q10 has imported this pull request. If you are a Meta employee, you can view this in D78466075.

1 similar comment

Contributor

facebook-github-bot commented Jul 19, 2025

@q10 has imported this pull request. If you are a Meta employee, you can view this in D78466075.


          Merge branch 'pytorch:main' into small-l-opt-c++

a599f18

Contributor

facebook-github-bot commented Jul 21, 2025

@q10 has imported this pull request. If you are a Meta employee, you can view this in D78466075.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels