Hugging Face's logo

ydmeira
/
beit-finetuned-pokemon

Transformers
PyTorch
TensorBoard
beit
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
beit-finetuned-pokemon / README.md
ydmeira's picture
ydmeira
update model card README.md
3ec96f7
preview code
|
raw
history blame
4.38 kB
metadata
license: apache-2.0
tags:
  - generated_from_trainer
model-index:
  - name: beit-finetuned-pokemon
    results: []

beit-finetuned-pokemon

This model is a fine-tuned version of microsoft/beit-base-finetuned-ade-640-640 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0426
  • Mean Accuracy: 0.9851
  • Mean Iou: 0.4926
  • Overall Accuracy: 0.9851
  • Per Category Accuracy: [nan, 0.9851295328900131]
  • Per Category Iou: [0.0, 0.9851295328900131]

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: 6e-05
  • train_batch_size: 2
  • eval_batch_size: 2
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 1

Training results

Training Loss Epoch Step Validation Loss Mean Accuracy Mean Iou Overall Accuracy Per Category Accuracy Per Category Iou
0.2845 0.05 250 0.1909 0.8750 0.4375 0.8750 [nan, 0.8750296526422883] [0.0, 0.8750296526422883]
0.103 0.11 500 0.1987 0.9048 0.4524 0.9048 [nan, 0.9047505435789185] [0.0, 0.9047505435789185]
0.091 0.16 750 0.2199 0.8935 0.4468 0.8935 [nan, 0.8935388953867466] [0.0, 0.8935388953867466]
0.0787 0.21 1000 0.0498 0.9832 0.4916 0.9832 [nan, 0.9832157481853218] [0.0, 0.9832157481853218]
0.0516 0.27 1250 0.0642 0.9767 0.4884 0.9767 [nan, 0.9767367885585835] [0.0, 0.9767367885585835]
0.051 0.32 1500 0.0907 0.9582 0.4791 0.9582 [nan, 0.9582013500039326] [0.0, 0.9582013500039326]
0.0518 0.37 1750 0.0813 0.9578 0.4789 0.9578 [nan, 0.9577983594953152] [0.0, 0.9577983594953152]
0.038 0.43 2000 0.0394 0.9875 0.4937 0.9875 [nan, 0.9874955917462267] [0.0, 0.9874955917462267]
0.0466 0.48 2250 0.0482 0.9831 0.4915 0.9831 [nan, 0.9830982793221819] [0.0, 0.9830982793221819]
0.054 0.53 2500 0.0568 0.9818 0.4909 0.9818 [nan, 0.9818346010498621] [0.0, 0.9818346010498621]
0.0356 0.59 2750 0.0330 0.9921 0.4961 0.9921 [nan, 0.9921038026421615] [0.0, 0.9921038026421615]
0.0292 0.64 3000 0.0364 0.9893 0.4947 0.9893 [nan, 0.9893293618878236] [0.0, 0.9893293618878236]
0.0252 0.69 3250 0.0607 0.9824 0.4912 0.9824 [nan, 0.9823825882221607] [0.0, 0.9823825882221607]
0.0286 0.75 3500 0.0526 0.9830 0.4915 0.9830 [nan, 0.9830357074898451] [0.0, 0.9830357074898451]
0.0297 0.8 3750 0.0403 0.9844 0.4922 0.9844 [nan, 0.9843719475221174] [0.0, 0.9843719475221174]
0.0257 0.85 4000 0.0478 0.9848 0.4924 0.9848 [nan, 0.9847944421751276] [0.0, 0.9847944421751276]
0.0271 0.91 4250 0.0340 0.9869 0.4935 0.9869 [nan, 0.9869270221516337] [0.0, 0.9869270221516337]
0.0235 0.96 4500 0.0426 0.9851 0.4926 0.9851 [nan, 0.9851295328900131] [0.0, 0.9851295328900131]

Framework versions

  • Transformers 4.21.2
  • Pytorch 1.12.1+cu113
  • Datasets 2.4.0
  • Tokenizers 0.12.1