regnet-y-064-Brain_Tumors_Image_Classification
This model is a fine-tuned version of facebook/regnet-y-064.
It achieves the following results on the evaluation set:
- Loss: 1.1561
- Accuracy: 0.8046
- Weighted f1: 0.7776
- Micro f1: 0.8046
- Macro f1: 0.7839
- Weighted recall: 0.8046
- Micro recall: 0.8046
- Macro recall: 0.7978
- Weighted precision: 0.8574
- Micro precision: 0.8046
- Macro precision: 0.8736
Model Description
Click here for the code that I used to create this model.This project is part of a comparison of seventeen (17) transformers.
Click here to see the README markdown file for the full project.Intended Uses & Limitations
This model is intended to demonstrate my ability to solve a complex problem using technology.
Training & Evaluation Data
Brain Tumor Image Classification DatasetSample Images
Class Distribution of Training Dataset
Class Distribution of Evaluation Dataset
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0002
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Weighted f1 | Micro f1 | Macro f1 | Weighted recall | Micro recall | Macro recall | Weighted precision | Micro precision | Macro precision |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1.288 | 1.0 | 180 | 1.3796 | 0.6548 | 0.5991 | 0.6548 | 0.5868 | 0.6548 | 0.6548 | 0.6176 | 0.8046 | 0.6548 | 0.8285 |
| 1.288 | 2.0 | 360 | 1.0964 | 0.7944 | 0.7687 | 0.7944 | 0.7755 | 0.7944 | 0.7944 | 0.7872 | 0.8555 | 0.7944 | 0.8727 |
| 0.1498 | 3.0 | 540 | 1.1561 | 0.8046 | 0.7776 | 0.8046 | 0.7839 | 0.8046 | 0.8046 | 0.7978 | 0.8574 | 0.8046 | 0.8736 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.0.0
- Datasets 2.11.0
- Tokenizers 0.13.3