microsoft/onnxruntime

Pull Requests Commits

Fix transformers optimizations for GPT-NeoX (#16743)

tianleiwu committed 2 years ago

Verified 95f96287

Work on eliminating Internal Compiler Error (#16741)

yuslepukhin committed 2 years ago

Verified e752cbe7

[DORT] Reduce global configs to make enabling dynamic shape easier (#16720)

wschin committed 2 years ago

Verified b71ebf91

[ROCm] fix kernel explorer GemmSoftmaxGemm test (#16735)

PeixuanZuo committed 2 years ago

Verified 9b549c64

[CANN EP] Fix Float16 support for CANN EP (#16733)

zhangsibo1129 committed 2 years ago

Verified 9ba5cdba

[ROCm] Generalize FastGeLU (#16623)

cloudhan committed 2 years ago

Verified 0cab7e1a

Add MAUI test app that can be used to test model loading and performance (#16658)

skottmckay committed 2 years ago

Verified ad90352a

Fix warning about uninitialized member (#16736)

cloudhan committed 2 years ago

Verified a45b8347

Upgrade old Python version in packaging pipeline (#16667)

edgchen1 committed 2 years ago

Verified df8843c4

Fix parameter naming that fails Doc generation. (#16717)

yuslepukhin committed 2 years ago

Verified b8c40b78

RunAsync C/CXX API (#16613)

RandySheriffH committed 2 years ago

Verified e1ca8ee6

Cuda: Decoder Masked Multihead Attention Q values get corrupted when using cross attention (#16721)

RyanUnderhill committed 2 years ago

Verified 2cf31a20

[WebNN EP] Make some types clearer (#16705)

Honry committed 2 years ago

Verified 2b7a94e6

atomicAdd returns previous value, not current value. (#16690)

RyanUnderhill committed 2 years ago

Verified 2ae041f3

[DORT] Enable aten::full by implementing extra logics to select EP (#16699)

wschin committed 2 years ago

Verified 44fd98eb

[ios] Enable `--use_extensions` with custom built iOS pod (#16711)

edgchen1 committed 2 years ago

Verified f236768d

Fix issue in constant-propagation inside function subgraph (#16330)

gramalingam committed 2 years ago

Verified 4faee2e4

[WebNN EP] Support several activation ops (#16693)

Honry committed 2 years ago

Verified ea43671e

[QNN EP] Fix error handling for Softmax/ReduceOps (#16700)

adrianlizarraga committed 2 years ago

Verified a189e76f

Add support for training apis to support custom ops (#16601)

baijumeswani committed 2 years ago

Verified 9889f0f5

[QNN EP] Add option to skip unit tests in the QNN NuGet packaging pipeline (#16164)

adrianlizarraga committed 2 years ago

Verified 19169afe

[C#, CPP] Introduce Float16/BFloat16 support and tests for C#, C++ (#16506)

yuslepukhin committed 2 years ago

Verified 853c4ff0

Add Stable Diffusion Benchmark on A100-PCIE-80GB (#16702)

tianleiwu committed 2 years ago

Verified 77b45c65

add more check to Web CI on cache restore (#16689)

mszhanyi committed 2 years ago

Verified 36b121d8

[ROCm] TunableOp: add hipBLASLt tuning logic (#16338)

mindest committed 2 years ago

Verified 810512c6

Fix CodeCoverage pipeline (#16684)

skottmckay committed 2 years ago

Verified a3fc04ba

[js/web] fix file size trim for wasm only .min.js (#16681)

fs-eire committed 2 years ago

Verified d1d65978

Call `lazy_reset_grad` in on-device training docs (#16696)

dannyfriar committed 2 years ago

Verified 5de2e2fb

Amx flag removal (#16527)

dipanjans12 committed 2 years ago

Verified a4616084

Triton Codegen for ORTModule (#15831)

Lafi7e committed 2 years ago

Verified c07a3b86

Newer Older