Gordon Rix

Thanks so much for the quick help. I am only using `consensus.hdf` files that have been made using `medaka 1.6.1` so I think that isn't the problem. I will work...

Ok so before I didn't run the mostly unmodified smolecule that uses `.fasta` files (but does not use poa) on the larger dataset because of how much time it would...

@ftostevin-ont @cjw85 In my efforts to isolate specific problematic reads, I am finding that isolating single reads doesn't seem to reproduce the issue. By inserting a try except clause in...

Based on a similar error message that appeared in [this issue on the PEPPER gihub](https://github.com/kishwarshafin/pepper/issues/103), I think this is a memory consumption issue. @cjw85 @ftostevin-ont any suggestions for smolecule settings...

Understood! Sorry, I thought I had made provided enough info for you to reproduce the issue, but if I did it was diffusely spread over a wall of text. Below...

https://github.com/gordonrix/medaka only two changes are adding the `smoleculeTest.fasta` file in the main folder, and commenting out 3 lines in `smolecule.py` edit: sorry, hold on, I didn't make a branch because...

I think this is the same bug as https://github.com/nanoporetech/medaka/issues/381, which does not include any modifications to `smolecule`, so probably not worth your time to look at my fork. Just please...

Thanks for looking into this! To be clear though, are you saying that GPU contexts are being created even when GPU resources are not available? I ask because I have...

Oh wonderful now I understand, ok I'll try that fix this week and let you know what I find

Closing this as I am no longer seeing this error, and it does seem to be more performant. However, I am still seeing smolecule fail on my virtual machine if...