Owen Vallis
Owen Vallis
@AminHP apologies for taking a minute to respond here and thanks for the PR. This seems like an interesting way to support larger batch sizes and I like that the...
Hello @AminHP, Regarding the shuffling that you mentioned, it shuffles the entire dataset before we start sampling batches, but batches will still add the samples per class in contiguous blocks...
This will be fixed in 0.18
Hi Hannes, we just need to update the setup and workflow tests. Looks like there are a few changes but not too bad. I've started a PR and I'll try...
Thanks for your patience, it's been a busy start to the year but I'll try and get the updates in this week.
Hi @Lorenzobattistela, looks like convnext wasn't introduced until TF v2.10. I wonder if we can do a TF version check for this? The other option is we could provide a...
Thanks @shinstra, I'll try and take a look into this. The lookup error may be caused by something in the result set returned by nmslib, but I'll have to dig...
This will be fixed in 0.18
Hi Tony, this is a missing implementation in the nmslib package. They don't currently support saving a brute force index. Having said that... we have a new release coming in...
Hi @ManafMukred, can you share some more details about the model architecture, hyperparameters, and dataset?