Hugging Face's logo

MingliangLiang3
/
DynamiCS-ViT-B-16-DataComp-DFN

Zero-Shot Image Classification
OpenCLIP
clip
vision-language-model
image-text-retrieval
research
long-tail
datacomp
Model card Files
xet
 Community
Mingliang Liang commited on
Commit
5811b69
·
verified ·
1 Parent(s): b590862

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -68,7 +68,7 @@ model.eval()
68
 
69
  ## Training Data
70
 
71
- The checkpoints were trained on a **DataComp-DFN** subset derived from DataComp-Large and filtered with DFN-2B. In the project paper, the accessible subset is described as approximately **130M** image-text pairs after accounting for unavailable or expired URLs.
72
 
73
  DynamiCS computes per-sample sampling probabilities from semantic image clusters built with:
74
 
 
68
 
69
  ## Training Data
70
 
71
+ The checkpoints were trained on a **DataComp-DFN** subset derived from DataComp-Large and filtered with DFN. In the project paper, the accessible subset is described as approximately **130M** image-text pairs after accounting for unavailable or expired URLs.
72
 
73
  DynamiCS computes per-sample sampling probabilities from semantic image clusters built with:
74