Edit model card

cvt-21-384-22k-finetuned

This model is a fine-tuned version of microsoft/cvt-21-384-22k on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0004
  • Accuracy: 1.0

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 2e-05
  • train_batch_size: 10
  • eval_batch_size: 4
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 30

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 1.0 90 0.1145 0.9777
No log 2.0 180 0.0646 0.9732
No log 3.0 270 0.0524 0.9821
No log 4.0 360 0.0144 0.9955
No log 5.0 450 0.0234 0.9911
0.3541 6.0 540 0.0189 0.9911
0.3541 7.0 630 0.0099 0.9955
0.3541 8.0 720 0.0253 0.9866
0.3541 9.0 810 0.0414 0.9866
0.3541 10.0 900 0.0034 1.0
0.3541 11.0 990 0.0099 0.9955
0.2485 12.0 1080 0.0004 1.0
0.2485 13.0 1170 0.0088 0.9955
0.2485 14.0 1260 0.0104 0.9955
0.2485 15.0 1350 0.0001 1.0
0.2485 16.0 1440 0.0098 0.9955
0.2229 17.0 1530 0.0002 1.0
0.2229 18.0 1620 0.0004 1.0
0.2229 19.0 1710 0.0002 1.0
0.2229 20.0 1800 0.0001 1.0
0.2229 21.0 1890 0.0005 1.0
0.2229 22.0 1980 0.0002 1.0
0.2192 23.0 2070 0.0006 1.0
0.2192 24.0 2160 0.0001 1.0
0.2192 25.0 2250 0.0013 1.0
0.2192 26.0 2340 0.0002 1.0
0.2192 27.0 2430 0.0002 1.0
0.211 28.0 2520 0.0012 1.0
0.211 29.0 2610 0.0013 1.0
0.211 30.0 2700 0.0004 1.0

Framework versions

  • Transformers 4.38.1
  • Pytorch 1.10.0+cu111
  • Datasets 2.17.1
  • Tokenizers 0.15.2
Downloads last month
3
Safetensors
Model size
31.3M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for b07611031/cvt-21-384-22k-finetuned

Finetuned
(6)
this model

Evaluation results