onnxruntime/onnxruntime/test
pengwa 16f5909f2d
Introduce shrunken gather operator (#15396)
### Introduce shrunken gather operator

Exist Gather operator schema won't guarantee output element count will
be smaller than input element count.
Actually, it is possible output element count >, =, or < input element
count.

For some cases we know for sure output element count MUST be <= input
element count, we will upstream those Gather operators to reduce compute
flops.

So this PR introduces an ShrunkenGather which explicitly guarantee
output count will be smaller than input count. The operator add
additional restriction on inputs, but still re-use existing Gather's
implementations plus input check during runtime.

This is a requirement for subsequent optimization (Draft PR:
https://github.com/microsoft/onnxruntime/pull/15401) we will do for
label sparsity and embedding sparsity.
2023-04-07 15:12:58 +08:00
..
api_tests_without_env
common Fix CUDA tests for Ampere cards, and bump layernorm tests opset version (#14761) 2023-03-17 09:30:01 -07:00
contrib_ops Adding FP16 Global Average Pool operator (#15324) 2023-04-05 09:38:02 -07:00
custom_op_registration
debug_node_inputs_outputs
eager remove device_id parameter out of ExecutionProvider::GetAllocator() (#14580) 2023-02-13 10:01:07 -08:00
framework Automatically enable tunable op usage for production models (#15156) 2023-04-06 13:52:47 +08:00
fuzzing
global_thread_pools [oneDNN] Improved thread handling (#13618) 2023-01-31 14:37:13 -08:00
ir
logging_apis Add logging APIs for custom operators (#14416) 2023-03-17 15:05:28 -07:00
mlas Add float16 Tanh support (#15048) 2023-03-16 18:57:38 -07:00
onnx Adopt linrtunner as the linting tool - take 2 (#15085) 2023-03-24 15:29:03 -07:00
opaque_api
optimizer [QNN EP] Support Resize's pytorch_half_pixel coordinate transformation mode on HTP (#15390) 2023-04-06 23:56:33 -07:00
perftest refactor 2023-03-15 13:23:56 +08:00
platform
proto
providers Introduce shrunken gather operator (#15396) 2023-04-07 15:12:58 +08:00
python Reuse QDQConv for ConvTranspose to generate the QDQ model (#15385) 2023-04-06 15:07:44 -07:00
quantization remove device_id parameter out of ExecutionProvider::GetAllocator() (#14580) 2023-02-13 10:01:07 -08:00
shared_lib Update the post-merge pipeline (#14965) 2023-03-29 13:12:07 -07:00
testdata Upstream reshape of merging batch/sequence (#15023) 2023-04-05 18:51:07 +08:00
unittest_main
util Automatically enable tunable op usage for production models (#15156) 2023-04-06 13:52:47 +08:00
wasm [js/web] allow unittest (onnxruntime_test_all) to run in browser (#14820) 2023-02-24 16:45:33 -08:00
win_getopt
xctest