The dataset could not be loaded because the splits use different data file formats, which is not supported. Read more about the splits configuration. Click for more details.

Couldn't infer the same data file format for all splits. Got {NamedSplit('train'): ('csv', {}), NamedSplit('test'): ('imagefolder', {})}

Error code:   FileFormatMismatchBetweenSplitsError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

LDW-CNet Training Logs & Evaluation Artefacts

This dataset repository hosts the training histories, evaluation reports, and figures for the LDW-CNet project. The underlying brain MRI images are NOT redistributed here — they come from the upstream Kaggle dataset [masoudnickparvar/brain-tumor-mri-dataset] (https://www.kaggle.com/datasets/masoudnickparvar/brain-tumor-mri-dataset).

Repository contents

training_logs/        Per-model CSV history (loss, F1, λ_wc, LR per epoch)
reports/              Evaluation CSVs (metrics, calibration, robustness, etc.)
figures/              All PNG figures from notebooks 3 and 4
MODEL_CARD.md         Top model's HF card (mirrored from model repo)
README.md             This file

Upstream dataset summary

Source: Kaggle — masoudnickparvar/brain-tumor-mri-dataset
Classes (4): glioma, meningioma, notumor, pituitary
Image modality: T1-weighted axial brain MRI slices, JPEG/PNG
Total images: ~7200

Class distribution

Class	Train	Test
glioma	1400	400
meningioma	1400	400
notumor	1400	400
pituitary	1400	400

How the data was used

Train/val split: 90% / 10% stratified random split from the Training/ directory, random_state=42
Test split: untouched Testing/ directory
Image size: 224×224
Normalisation: ImageNet mean/std

Reproducing the artefacts

The complete pipeline lives in four notebooks (see model repo for the code itself):

01_Setup_and_Data — environment, config, exploratory analysis
02_Model_Architecture — model class definitions
03_Training — full training loop with all 5 model variants
04_Evaluation — this notebook's outputs

Important note on clinical use

This dataset card describes research artefacts. None of these models are medical devices. Do not use them for clinical decision making.

Citations

For the underlying images, cite the upstream dataset author. For these artefacts, cite the model card in the companion model repo.

License

MIT (for the artefacts in this repo). The upstream image dataset retains its original license — see the Kaggle dataset page.

Downloads last month: 57

Total file size:

11.1 MB