cvt-21-384-22k-finetuned-PinnatelyCompound

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: 0.0064
  • Accuracy: 0.9955

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: 40
  • eval_batch_size: 16
  • 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 23 0.2918 0.8973
No log 2.0 46 0.1656 0.9330
No log 3.0 69 0.0529 0.9821
No log 4.0 92 0.0144 0.9955
No log 5.0 115 0.0266 0.9911
No log 6.0 138 0.0244 0.9955
No log 7.0 161 0.0144 0.9955
No log 8.0 184 0.0154 0.9955
No log 9.0 207 0.0188 0.9911
No log 10.0 230 0.0094 0.9955
No log 11.0 253 0.0055 1.0
No log 12.0 276 0.0026 1.0
No log 13.0 299 0.0057 1.0
No log 14.0 322 0.0079 0.9955
No log 15.0 345 0.0026 1.0
No log 16.0 368 0.0017 1.0
No log 17.0 391 0.0044 0.9955
No log 18.0 414 0.0038 1.0
No log 19.0 437 0.0120 0.9911
No log 20.0 460 0.0005 1.0
No log 21.0 483 0.0019 1.0
0.2553 22.0 506 0.0020 1.0
0.2553 23.0 529 0.0026 1.0
0.2553 24.0 552 0.0053 0.9955
0.2553 25.0 575 0.0009 1.0
0.2553 26.0 598 0.0008 1.0
0.2553 27.0 621 0.0016 1.0
0.2553 28.0 644 0.0010 1.0
0.2553 29.0 667 0.0008 1.0
0.2553 30.0 690 0.0064 0.9955

Framework versions

  • Transformers 4.38.1
  • Pytorch 1.10.0+cu111
  • Datasets 2.17.1
  • Tokenizers 0.15.2
Downloads last month
17
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-PinnatelyCompound

Finetuned
(6)
this model

Evaluation results