39c865b3aa
Set the acceptable error range based on both absolute tolerance and relative tolerance. Currently, absolute tolerance is set to 1e-5 for FP32 and 5 epsilon (~5e-3) for FP16 relaxed computation. The relative tolerance is set to 5ULP of the corresponding precision. Add a TODO mark for potential future adjustment on error limit based on testing. Bug: 111768023 Test: none Change-Id: Idedcec3e09fd7de9696811b93c81d0f180e896ef |
||
---|---|---|
.. | ||
1.0 | ||
1.1 | ||
README |
NeuralNetworks sample driver implementation is located at frameworks/ml/nn/driver/sample.