Jan-Frederik Schulte
Jan-Frederik Schulte
Fair point. The other thing brought up at the meeting was if we can rebalance the batching of tests somehow? Right now some of them take a few minutes while...
Looks like the new tests actually have one failure in the latest run: ``` FAILED test_keras_api.py::test_conv2d[io_parallel-Vitis-Resource-same-channels_last] - AssertionError: BestLatency: expected 53.0, got 52.0 (tolerance=0%) ``` As for the other feedback,...
It's not available yet, we expect a pull request integrating this code in the next weeks.
Is there anything preventing us from merging this?
@calad0i Do you think this needs to be included explicitly also in the keras v3 parser, or should we let the fallback take care of it?
As discussed in today's dev meeting, we are merging this PR now with the understanding that it might be good to generalize in the future as to not have separate...
Generally this looks good to me, comments are minor. I'll wait until some things are merged that should fix some tests failures and then run the CI.
Test failures unrelated, this is ready for merge.
We support Vitis HLS 2024.1. 2024.2 will probably also work, but it is not fully validated. We do not support using hls4ml on Windows at all, see https://fastmachinelearning.org/hls4ml/intro/status.html.
Hi! Thanks for reporting these issues. The softmax issues are indeed a bug that we overlooked because softmax activation was missing from our tests. This is fixed in https://github.com/fastmachinelearning/hls4ml/pull/1086. For...