vit-msn-small-wbc-classifier-0316-cleaned-dataset-10

This model is a fine-tuned version of facebook/vit-msn-small on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3538
  • Accuracy: 0.8935

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: 5e-05
  • train_batch_size: 64
  • eval_batch_size: 64
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 256
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 25

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.5193 1.0 16 0.6823 0.7945
0.5339 2.0 32 0.4553 0.8438
0.4778 3.0 48 0.4525 0.8478
0.4253 4.0 64 0.4077 0.8473
0.4086 5.0 80 0.4218 0.8575
0.3673 6.0 96 0.4002 0.8693
0.3275 7.0 112 0.3302 0.8773
0.3231 8.0 128 0.3672 0.8803
0.302 9.0 144 0.3363 0.8900
0.3122 10.0 160 0.3284 0.8843
0.2686 11.0 176 0.3317 0.8874
0.2786 12.0 192 0.3660 0.8883
0.2338 13.0 208 0.3520 0.8834
0.2466 14.0 224 0.3414 0.8896
0.2296 15.0 240 0.3531 0.8874
0.1961 16.0 256 0.3844 0.8847
0.2056 17.0 272 0.3705 0.8900
0.197 18.0 288 0.3538 0.8935
0.1748 19.0 304 0.3717 0.8887
0.1807 20.0 320 0.4075 0.8843
0.177 21.0 336 0.3881 0.8830
0.1433 22.0 352 0.4014 0.8856
0.1522 23.0 368 0.3918 0.8874
0.1322 24.0 384 0.4199 0.8905
0.1396 25.0 400 0.4142 0.8896

Framework versions

  • Transformers 4.44.2
  • Pytorch 2.4.1+cu121
  • Datasets 3.2.0
  • Tokenizers 0.19.1
Downloads last month
133
Safetensors
Model size
21.7M 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 Melo1512/vit-msn-small-wbc-classifier-0316-cleaned-dataset-10

Finetuned
(25)
this model

Evaluation results