JoshuaKelleyDs's picture
Update README.md
4b5f632 verified
metadata
license: other
base_model: google/mobilenet_v2_1.0_224
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: doodle_mobilenet
    results: []

doodle_mobilenet

This model is a fine-tuned version of google/mobilenet_v2_1.0_224 on the
Quick, Draw! small dataset

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: 0.0008
  • train_batch_size: 512
  • eval_batch_size: 512
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 10
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.4546 0.5688 5000 1.4383 0.6474
1.3759 1.1377 10000 1.3850 0.6610
1.3508 1.7065 15000 1.3163 0.6737
1.294 2.2753 20000 1.2832 0.6829
1.2811 2.8441 25000 1.2581 0.6881
1.2331 3.4130 30000 1.2387 0.6926
1.2276 3.9818 35000 1.2227 0.6978
1.1964 4.5506 40000 1.2196 0.6990
1.1498 5.1195 45000 1.1994 0.7036
1.1548 5.6883 50000 1.1900 0.7052
1.1232 6.2571 55000 1.1831 0.7075
1.1264 6.8259 60000 1.1695 0.7100
1.0896 7.3948 65000 1.1584 0.7128
1.0917 7.9636 70000 1.1535 0.7155
1.0654 8.5324 75000 1.1545 0.7144
1.0395 9.1013 80000 1.1471 0.7169
1.0383 9.6701 85000 1.1722 0.7136

Framework versions

  • Transformers 4.40.0
  • Pytorch 2.2.1
  • Datasets 2.19.0
  • Tokenizers 0.19.1