Gitiles
Code Review
Sign In
gerrit-public.fairphone.software
/
platform
/
test
/
mlts
/
benchmark
/
refs/heads/odm/dev/target/13/fp5
/
src
/
com
/
android
/
nn
/
benchmark
/
core
32f40e1
Fix memory leak
by Slava Shklyaev
· 3 years, 10 months ago
2aaa3f8
Refactor code to prepare for parallel tests
by Stefano Galarraga
· 4 years, 4 months ago
5e82b1b
Adds a method to check if there's an NNAPI accelerator available.
by Viet Dang
· 4 years, 8 months ago
f7302c1
Extract NN API Benchmarks into an app that can easily be run by dogfooders.
by Andy Dyer-Smith
· 5 years ago
226e8f1
Fix scoring test for unsupported models and models without evaluators.
by Viet Dang
· 5 years ago
851f4b6
Add support for restricting NNAPI to one device
by Slava Shklyaev
· 5 years ago
d0dddca
Add missing test name to CSV output.
by Przemyslaw Szczepaniak
· 5 years ago
3361681
Adds some getters to BenchmarkResult class.
by Viet Dang
· 5 years ago
29e49bf
Removes the constraint that inputSize needs to be 4-element array.
by Viet Dang
· 5 years ago
1f0f440
Add a benchmark library to support the CTS test.
by Viet Dang
· 5 years ago
9aaf234
Separates UnsupportedSdkException from BenchmarkException.
by Viet Dang
· 5 years ago
ffec38a
Adds a minSdkVersion property to test model config.
by Viet Dang
· 6 years ago
ab62f60
Converts benchmark JNI lib to NDK (*.mk) build to support older SDKs.
by Viet Dang
· 6 years ago
71e2411
Fix NNTestBase::runInferenceOnce()
by Viet Dang
· 6 years ago
047c632
Support models with multiple outputs.
by Przemyslaw Szczepaniak
· 6 years ago
19f760b
Fix for NPE due to null BenchmarkResult#mValidationErrors array.
by Przemyslaw Szczepaniak
· 6 years ago
80758ca
Fix for AssetManager.list() not working on some devices.
by Przemyslaw Szczepaniak
· 6 years ago
507d3f6
Expose validation errors in score summary.
by Przemyslaw Szczepaniak
· 6 years ago
456f75a
Merge "Renames AssetUtils to IOUtils and moves readFloats() there."
by Viet Dang
· 6 years ago
3b903b9
Fix input dataset size calculation
by Przemyslaw Szczepaniak
· 6 years ago
bf3f041
Run whole dataset in scoring benchmark.
by Przemyslaw Szczepaniak
· 6 years ago
9429af4
Generate results summary HTML page.
by Przemyslaw Szczepaniak
· 6 years ago
9cb72d9
Renames AssetUtils to IOUtils and moves readFloats() there.
by Viet Dang
· 6 years ago
4b63f23
Adds validation for the TTS model.
by Viet Dang
· 6 years ago
b8cf224
Makes evaluator field an object and refactors asset reading logic.
by Viet Dang
· 6 years ago
94771d0
Utility for dumping intermediate tensors.
by Przemyslaw Szczepaniak
· 6 years ago
26b3d80
Moves evaluator field one level up in the model config loader.
by Viet Dang
· 6 years ago
bcdd259
Switch the benchmark app to using the new NNAPI Delegate.
by Viet Dang
· 6 years ago
04f845b
Move dataSize field one level up in test list parser.
by Lev Proleev
· 6 years ago
7cfcee7
Support datasets and top-k evaluation in mlts
by Mika Raento
· 6 years ago
6063374
Adds support for input/output sequence to the benchmark.
by Viet Dang
· 6 years ago
c6b1aab
Support TFLite-only execution in benchmark
by Mika Raento
· 6 years ago
e1059cb
Add benchmark run flags
by Przemyslaw Szczepaniak
· 6 years ago
ddf02b5
Add max single element error to benchmark output
by Mika Raento
· 6 years ago
1fd8036
Extract 'core' benchmark classes into a seperate package.
by Przemyslaw Szczepaniak
· 6 years ago