EntityNet
Collection
The models of our publication: Using Knowledge Graphs to harvest datasets for efficient CLIP model training
•
7 items
•
Updated
A CLIP (Contrastive Language-Image Pre-training) model finetuned on EntityNet-33M. The base model is ViT-B-16/datacomp_xl_s13b_b90k.
See the project page for the paper, code, usage examples, metrics, etc.
The model has seen ~13B images at a batch size of 90k during pretraining, and ~0.6B images at a batch size of 32k during finetuning.
- Downloads last month
- 3
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support