microsoft/onnxruntime

Pull Requests Commits

optional multi-threading

RandyShuai committed 5 years ago

94dd59bd

RandyShuai committed 5 years ago

371a4be1

RandyShuai committed 5 years ago

ff4f6a5b

RandyShuai committed 5 years ago

7249161c

RandyShuai committed 5 years ago

31e76d31

implement multi-threading expand on cpu

RandyShuai committed 5 years ago

83aed528

Fix kernel registrations for Equal, Greater and Less (#5730)

skottmckay committed 5 years ago

Verified c0c9ab4d

Add tag types for Ort::Float16_t and Ort:Bfloat16_t structs (#5716)

yuslepukhin committed 5 years ago

Verified 2bf5046d

Add GPU kernels for ROCm EP (#5655)

weixingzhang committed 5 years ago

Verified fff85a6a

Skip failing x86 winml tests and update testData environment variable path mechanism (#5719)

Ryan Lai committed 5 years ago

Verified 697e8faa

added missing flag ORT_TENSORRT_DUMP_SUBGRAPHS (#5724)

joba01 committed 5 years ago

Verified 9ec6da1e

added missing flag ORT_TENSORRT_DUMP_SUBGRAPHS (#5724)

joba01 committed 5 years ago

Verified 6f6dd0b8

Tensorrt perf tool (#5436)

chilo-ms committed 5 years ago

Verified 92292de1

Revert saving optimized model as external data (#5690)

wangyems committed 5 years ago

Verified 95e6da79

Nuget packaging no omp (#5666)

RandySheriffH committed 5 years ago

Verified 71f90e08

Add option to allow quantize_input() use input_qtype for initializers. (#5721)

zhanghuanrong committed 5 years ago

Verified 77b1eea9

update jetson build instructions (#5725)

jywu-msft committed 5 years ago

Verified f666c3d7

Prepacking in Gemm with merged logic for Matmul and Gemm on PackingB. (#5693)

zhanghuanrong committed 5 years ago

Verified 24016a51

Add link to survey to README (#5685)

natke committed 5 years ago

Verified 479ed740

Modify logic to determine OV Version (#5701)

MaajidKhan committed 5 years ago

Verified d6f9cc18

[Java] Fixes an error allocating large direct byte buffers during OnnxTensor creation (#5619)

Craigacp committed 5 years ago

Verified d1d82065

Register opset13 flatten, LRN for cuda. (#5694)

Pranav Sharma committed 5 years ago

Verified 28197b14

Partition full graph one execution provider at a time (#5635)

skottmckay committed 5 years ago

Verified 11fe6834

Implement reduce_matrix_columns() to optimize ReduceSum (#5639)

edgchen1 committed 5 years ago

Verified 858040fa

[TensorRT EP] Remove cudaDeviceSynchronize and use cudaAllocator for scratch buffers (#5714)

jywu-msft committed 5 years ago

Verified c46515cd

Remove redundant const_cast (#5705)

yuslepukhin committed 5 years ago

Verified fd9d0c4e

Add static CRT DLLs to Nuget package (#5661)

tiagoshibata committed 5 years ago

Verified 9e68e984

Unbreak microbenchmark build (#5710)

tlh20 committed 5 years ago

Verified ff23083d

Calibrate float tensor only (#5704)

yufenglee committed 5 years ago

Verified 5c4543e1

The IndexedSubGraph is used to create the Function body, but after that is invalid as the nodes it referred to have been removed from the main Graph. As such there's no need to store it in the FunctionImpl instance. (#5669)

skottmckay committed 5 years ago

Verified 2127a229

Older