Hugging Face's logo

kariver
/
git-base-food

Image-Text-to-Text
Transformers
PyTorch
git
Generated from Trainer
Inference Endpoints
Model card Files Community
git-base-food / README.md
kariver's picture
kariver
End of training
f501cca
preview code
|
raw
history blame
2.18 kB
metadata
license: mit
base_model: microsoft/git-base
tags:
  - generated_from_trainer
datasets:
  - imagefolder
model-index:
  - name: git-base-food
    results: []

git-base-food

This model is a fine-tuned version of microsoft/git-base on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0621
  • Wer Score: 4.7174

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: 8
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 16
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 20

Training results

Training Loss Epoch Step Validation Loss Wer Score
No log 1.48 20 7.3896 132.1522
No log 2.96 40 5.5234 1.1413
No log 4.44 60 3.7317 1.1196
No log 5.93 80 2.1477 1.1304
No log 7.41 100 0.9848 1.1087
No log 8.89 120 0.3860 1.1304
No log 10.37 140 0.1720 1.1087
No log 11.85 160 0.1022 1.1196
No log 13.33 180 0.0745 1.0543
No log 14.81 200 0.0665 2.9239
No log 16.3 220 0.0631 2.7174
No log 17.78 240 0.0622 2.5326
No log 19.26 260 0.0621 4.7174

Framework versions

  • Transformers 4.34.0
  • Pytorch 2.0.1+cu118
  • Datasets 2.14.5
  • Tokenizers 0.14.1