| ---
|
| license: cc-by-4.0
|
| tags:
|
| - balanced
|
| - csv
|
| - heavy
|
| - image-depth
|
| - kfold-5
|
| - nlp-sentiment
|
| - progressive
|
| - strict
|
| - weak-supervision
|
| ---
|
|
|
| # clean.py
|
|
|
| ## Dataset Summary
|
|
|
| A **nlp sentiment** dataset with **image depth** modality, stored in **csv** format.
|
|
|
| ## Preprocessing & Augmentation
|
|
|
| - **Preprocessing**: progressive
|
| - **Augmentation**: heavy
|
|
|
| ## Splits & Sampling
|
|
|
| - **Split strategy**: kfold 5
|
| - **Sampling**: balanced
|
|
|
| ## Quality & Labeling
|
|
|
| - **Quality filtering**: strict
|
| - **Labeling**: weak supervision
|
|
|
| ## Files
|
|
|
| - `clean.py` — main artifact of this repository
|
|
|
| ## License
|
|
|
| See the license field above.
|
|
|
