microsoft/onnxruntime

Pull Requests Commits

Enable type reduction in EyeLike, Mod, random.cc CPU kernels. (#6960)

edgchen1 committed 4 years ago

Verified d5ed3e7f

fix stream sync issue (#6954)

tianleiwu committed 4 years ago

Verified 89916fdb

Update baseline due to loss scale fix (#6948)

wschin committed 4 years ago

Verified bdaea1d9

Fix broken link in server usage and remove absolute path from dockerfiles readme (#6926)

0xRaduan committed 4 years ago

Verified 743a93fa

Add GPU support for DNNL endpoint (#6741)

georgen117 committed 4 years ago

Verified ba51774a

Support parsing an array of values stored as an attribute in a custom op (#6878)

hariharans29 committed 4 years ago

Verified c8e2e319

Enable build with bitcode for iOS (#6905)

guoyu-wang committed 4 years ago

Verified e64eff1f

Rename op kernel type control 'supported types' to 'default types'. (#6886)

edgchen1 committed 4 years ago

Verified 73fe1f2d

Only set _native folder for Microsoft.AI.MachineLearning package (#6939)

smk2007 committed 4 years ago

Verified 67c67408

MLAS: workaround LLVM x86 assembler (#6922)

tracysh committed 4 years ago

Verified bc276521

Add tests of Attention and QAttention for pruned model (#6914)

tianleiwu committed 4 years ago

Verified b89f52c2

Fix broken Java API link (#6826)

dennyac committed 4 years ago

Verified f2f60eed

Enable type reduction for Clip, MaxPool, and Pad CPU kernels. (#6918)

edgchen1 committed 4 years ago

Verified 15d81fb6

Support required types when excluding typed registrations (#6871)

edgchen1 committed 4 years ago

Verified b6c4a7ac

Fix loss scaling when running ORTTrainer with BERT under mixed-precision mode (#6932)

wschin committed 4 years ago

Verified de6e66f3

update Readme (#6903)

jywu-msft committed 4 years ago

Verified 601e04fb

Ability to fuse non-square (pruned) attention weights for BERT-like models (#6850)

mfuntowicz committed 4 years ago

Verified 9126faa3

move pipeline file and change relative path (#6882)

RandySheriffH committed 4 years ago

Verified f986ffcb

No such file or directory with --use_external_data_form and int8 (#6867)

rdzotz committed 4 years ago

Verified 107c9672

Configure session thread pool spinning preference (#6895)

RandySheriffH committed 4 years ago

Verified 679718b1

Save output tensors in bert_test_data tool (#6872)

tianleiwu committed 4 years ago

Verified 8f1786d5

Fix app packaging in UWP (#6804)

tiagoshibata committed 4 years ago

Verified fa8d1b44

Revert Gather Grad optimization in PR 6381 targeted for Rocm (#6880)

Suffian Khan committed 4 years ago

Verified 7915b670

Add check that the first 2 Loop subgraph inputs have an shape (could be explicit or inferred) as we need to know the rank the subgraph expects. Other inputs to the subgraph are more opaque so we can just pass them through. (#6891)

skottmckay committed 4 years ago

Verified 54cdb6af

Move constants from heap to stack to avoid randomness on cudnn function (#6869)

RandySheriffH committed 4 years ago

Verified d01006fc

switch to OrtMutex

RandyShuai committed 4 years ago

324dd2fa

Workaround for HTTP Error 403: Forbidden for MNIST dataset (#6885)

baijumeswani committed 4 years ago

Verified ed1883a9

switch to shared_timed_mutex

RandyShuai committed 4 years ago

6e60711b

[NNAPI EP] Add per-tensor u8s8 support for Qlinear[Conv/MatMul] (#6818)

guoyu-wang committed 4 years ago

Verified fedb6842

switch to shared_lock

RandyShuai committed 4 years ago

81f24ea6

Newer Older