Niels Ole Salscheider
Niels Ole Salscheider
We are also interested in this. Maybe @codemzs can comment on the plans since he reverted it?
Sorry, I haven't worked on this for a long time. I think this PR is a good starting point to base your implementation on, but back then I could not...
I think you can get the path using $. If that path is added to the PATH variable then the tests should run.
We would be really interested in that feature.
I like the proposal and think that we should do it that way for 1.0.
I am seeing the same, but the buffer size is already set to the smaller one.
This would be really helpful for deploying trained models in production. Many embedded accelerators require you to use the vendor graph compilers and inference APIs. And even if you could...