Davide Baldelli
Davide Baldelli
## New Models Added support for: - ViT base and large - ConvNeXt base, large and xxl - New ResNets - All [CLIP](https://github.com/openai/CLIP) and [OpenCLIP](https://github.com/mlfoundations/open_clip) models *For CLIP and OpenCLIP...
I am not understanding the conceptual usefulness of masking out the prompt. I have seen that there is a comment in scripts/prepare_alpaca.py that says: `mask_inputs: bool = False, # as...
In order to make the test more versatile I believe it would be useful to implement **shorter** versions of the test. In particular, a shared limitation of existing psychometric measures...