[RFC] UMA: Universal Modular Accelerator Interface

[slai-nick]

Hi, I am looking for some help too with UMA, I couldn’t get a keras model to compile for the vanilla accelerator. There’s a potential bug as well so it might be worth checking it out.

UMA & from_keras: putting it all together

Hi, I’ve been trying to go beyond the vanilla example trying to compile a simple keras model consisting of one Conv2D - I couldn’t get it to work. Here is the code for run_model.py (EDIT:: edited to match the following edit comments): from tvm.micro.testing.aot_test_utils import AOT_DEFAULT_RUNNER import tvm from tvm import relay from backend import VanillaAcceleratorBackend from tvm.relay import transform from collections import OrderedDict import numpy as np from tvm.testing.aot import ( …