Create readme.txt
Browse files
my_model/dataset/readme.txt
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
Directory Overview: This directory contains all the OK-VQA dataset raw files excluding the images in addition to the `dataset_processor.py` module.
|
| 2 |
+
|
| 3 |
+
dataset_processor.py module:
|
| 4 |
+
The `dataset_processor.py` module is designed to handle the OK-VQA dataset, providing functionality to load data from JSON files, merge question and annotation data, process answers, and save the processed data to CSV files.
|
| 5 |
+
|
| 6 |
+
Notes:
|
| 7 |
+
Ensure the necessary dependencies are installed (json, contractions, csv, pandas).
|
| 8 |
+
The module supports loading data from JSON files, merging question and annotation data, processing answers, and saving the processed data to CSV files.
|
| 9 |
+
The process_okvqa_dataset function provides an easy-to-use interface for processing the OK-VQA dataset by specifying the file paths for questions and annotations.
|