To run the demo, make sure to build the project, activate the environment, and set the appropriate environment variables. For more information, refer installation and build guide.
When running this model on N300 or T3000, make sure to place dispatch on ethernet cores with export WH_ARCH_YAML=wormhole_b0_80_arch_eth_dispatch.yaml for optimal performance
To run UNet Shallow for multiple iterations on single-chip at the best performance:
pytest --disable-warnings models/experimental/functional_unet/tests/test_unet_trace.py::test_unet_trace_2cq_same_ioTo run UNet Shallow for multiple iterations on N300 and T3000 at the best performance:
pytest --disable-warnings models/experimental/functional_unet/tests/test_unet_trace.py::test_unet_trace_2cq_multi_deviceUse pytest models/experimental/functional_unet/tests/test_unet_model.py to run the functional UNet Shallow model on a single-chip.
- N150
- N300
- T3K
- Inputs by default are random data.
- Weights by default are random data. We apply optimizations to tensors (such as batch folding) and as such, we modify the weights to ensure we don't achieve inf values.