Cao E

icon indicating an email [email protected]

icon company intel

Results 59 comments of Cao E

[Inductor][CPP] Leverage full bits for bf16/fp16 vectorization

@pytorchbot merge

[Inductor][CPP] Leverage full bits for bf16/fp16 vectorization

@pytorchbot merge

[Inductor][CPP] Leverage full bits for bf16/fp16 vectorization

@pytorchbot merge

[Inductor][CPP] Leverage full bits for bf16/fp16 vectorization

@pytorchbot merge

Add oneDNN support for Half LSTM on CPU

@yanbing-j @jgong5 Could you please review this PR again ? Thanks.

Add oneDNN support for Half LSTM on CPU

@yanbing-j @jgong5 Could you please review this PR again ? Thanks.

Add oneDNN support for Half LSTM on CPU

@pytorchbot merge

Add oneDNN support for Half LSTM on CPU

@pytorchbot merge

Add specializations for VecMaskLoad and VecMaskCast

> How are we testing this? Are there existing tests or do we need to add some? Hi @jansel. We add test cases for mask related functions in vec_test_all_types.cpp.

Add specializations for VecMaskLoad and VecMaskCast

Hi @jansel Could you please help review this PR again ? Thank you.