Owen Vallis
Owen Vallis
Hi, I don't have an M1 to test on at the moment, but I was able to install on a 2019 MacBookPro OS 12.0.1 using the following: ``` python3 -m...
Hi, I moved the tensorflow dep to extras_require. Seems like this is a common pattern to avoid tensorflow install issues. tensorflow_similarity 0.14.9 contains the fix. Please let me know if...
That's correct, I added the SplitValidationLoss to monitor the matching classification performance on the known vs the unknown classes to get a sense of how well an embedding is generalizing...
Rename metrics to something like K-Precision to make it clear these are our match metrics.
I've run into similar issues with interleave and have also recently come to feel that deterministic needs to be set to True in order to ensure the correct number of...
I haven't tested this yet, but I think one potential solution could be something like the following: ``` NUM_SHARDS_PER_CLASS = # num_images_in_class // num_examples_per_class def create_records(...): # Here we materialize...
Hi @kechan, I'll try and setup a working example later today, and apologies for any confusing var names above. To clarify, I'm hoping the above approach will enable use to:...
Hi Kapil, I just pushed some updates to the TFRecordDatasetSampler and they should be available on pypi now. The sampler should now work as expected if each class is a...
See #151 for the changes.
quick update to say I'm going to try and look at this later this week.