strisunshinewentingwang/pytorch

Pull Requests Commits

Add `pad_sequence` as a native function (#57868)

eqy committed 3 years ago

18edb77a

make ddp logging api to be private (#57999)

zhaojuanmao committed 3 years ago

ac44569b

[DataLoader] Remove redundant len >= 0 (#57951)

ejguan committed 3 years ago

e0539b0b

Enable cusolver potrf batched for Cholesky decomposition when cuda >= 11.3 (#57788)

xwang233 committed 3 years ago

7faac089

enable static graph training in DDP (#55248)

zhaojuanmao committed 3 years ago

ea421fb2

OpInfo: chunk (#57935)

kshitij12345 committed 3 years ago

502eb664

refactor multi_head_attention_forward (#56674)

Basil Hosmer committed 3 years ago

90f05c00

Add dot implementation for BFloat16 on CUDA (#57903)

imaginary-person committed 3 years ago

4fb8676c

Enable BFloat16 for `logaddexp` & `logaddexp2` on CUDA (#57908)

imaginary-person committed 3 years ago

067147ac

Enable geometric ops, exp2, expm1, rsqrt & erfc for BFloat16 on CUDA (#57913)

imaginary-person committed 3 years ago

fa318911

[CUDA graphs] Sync after replay (#57556)

mcarilli committed 3 years ago

dbedb1fa

[iOS GPU][perf][5/n] Replace std:vector with IntArrayRef and SmallVector (#57668)

xta0 committed 3 years ago

565550d8

[fbgemm] fix bug handling bias in rowwise quantization of FC (#58022)

jspark1105 committed 3 years ago

dc55ab3f

Update docs to mention CUDA support for Future (#50048)

lw committed 3 years ago

3e46d6c9

combine consecutive layes on the same device (#55973)

mrzzd committed 3 years ago

9e94921a

Use RPC context streams to cover serde ops (#57926)

mrshenli committed 3 years ago

cf7a0e5a

[iOS GPU][Perf][4/n] Reuse the same command buffer when copying results to CPU (#57667)

xta0 committed 3 years ago

0d564904

Deprecate chain_matmul (#57735)

lezcano committed 3 years ago

43f6deb6

Deprecate matrix_rank (#57734)

lezcano committed 3 years ago

7707efed

Deprecate torch.eig (#57727)

lezcano committed 3 years ago

415ae54c

Support mix of int32 and int64 offsets/indices for EmbeddingBag and its variants (#55189)

zyan0 committed 3 years ago

ee48bd08

TST Migrates some of test_nn.py from assertEqualIgnoreTypes to assertEqual (#57642)

thomasjpfan committed 3 years ago

3ec16035

Deprecate QR (#57745)

lezcano committed 3 years ago

24087d07

Deprecate torch.cholesky (#57725)

lezcano committed 3 years ago

4fef1c1d

[iOS GPU][Perf][3/n] Cache the compuation pipeline state object (#57666)

xta0 committed 3 years ago

f3e014f3

[fx ir] Handle the case when output consumes get_attr directly (#57844)

khabinov committed 3 years ago

36a22967

Alias det, slogdet, matrix_power, inverse, pinverse (#57821)

lezcano committed 3 years ago

a93314de

Deprecate torch.lstsq (#57743)

lezcano committed 3 years ago

ba84c911

[nccl] log rank when communicator is aborted (#57974)

rohan-varma committed 3 years ago

5840c8cf

OpInfo: select (#57731)

kshitij12345 committed 3 years ago

a0d686c9

Newer Older