This collection contains out-of-distribution (OOD) benchmarks used to evaluate BIOMEDICA models (CVPR 2025).
BIOMEDICA
community
AI & ML interests
None defined yet.
Recent Activity
BMC-SmolVLM1 is a family of lightweight biomedical vision-language models, ranging from 256 million to 2.2 billion parameters, based on SmolVLM.
A collection of prefiltered subsets in Webdataset format based on the BIOMEDICA 24M dataset.
This collection contains out-of-distribution (OOD) benchmarks used to evaluate BIOMEDICA models (CVPR 2025).
BMC-CLIP models are a state of the art Biomedical vision-lange embedding models.
BMC-SmolVLM1 is a family of lightweight biomedical vision-language models, ranging from 256 million to 2.2 billion parameters, based on SmolVLM.
A collection of prefiltered subsets in Parquet format based on the BIOMEDICA 24M dataset.
A collection of prefiltered subsets in Webdataset format based on the BIOMEDICA 24M dataset.