Model Card

Training Details

Adam Optimizer with a constant learning rate 1e-5 for 4000 steps training (batch_size=128). Only the vision encoder is fine-tuned.

alt text

Evaluation Results

Validation set accuracy:

  • pre-trained: 0.6803921461105347
  • fine-tuned: 0.8980392217636108

Test set accuracy:

  • pre-trained: 0.6644982695579529
  • fine-tuned: 0.8855098485946655
Downloads last month
533
Safetensors
Model size
87.5M params
Tensor type
F32
·
Inference API
Unable to determine this model's library. Check the docs .

Model tree for tanganke/clip-vit-base-patch32_oxford_flowers102

Finetuned
(56)
this model

Dataset used to train tanganke/clip-vit-base-patch32_oxford_flowers102