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.

Evaluation Results

Test set accuracy:

  • pre-trained: 0.8236435651779175
  • fine-tuned: 0.8840792179107666
Downloads last month
133
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_food101

Finetuned
(54)
this model

Dataset used to train tanganke/clip-vit-base-patch32_food101