datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
m3-retrieve-csrm3-retrieve-csr-cardiology-1kImage-ColorizationCSR-precompute-embedsPrecompute Imagenet1k Embeds for Beyond Matryoshka: Revisiting Sparse Coding for Adaptive Representation
For FF2048 backbones (same backbone weights with MRL), you can find the embeddings in FF2048_RN50_Embeds folder.
For SoTA backbones, you can find the embeddings in SoTA_RN50_Embeds folder.
For example, you can get embeds and corresponding label using:
import numpy as np
comp_data = np.load('FF2048_RN50_Embeds/1K_val_ff2048.npz')
embeds = comp_data['data']
labels = comp_data['label']
fnt-colpali-esrs-csrd