Upload README.md with huggingface_hub

README.md

@@ -1,19 +1,23 @@
 ---
-dataset_info:
-  features:
-  - name: image
-    dtype: image
-  - name: caption
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 90046462
-    num_examples: 2000
-  download_size: 103236915
-  dataset_size: 90046462
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
+language: en
+license: mit
+tags:
+- lighting
+- image-captioning
+- dataset
+- vision-language
 ---
+
+This dataset contains light-effect images and paired English captions.
+
+## Contents
+
+- `train`: prepared split with embedded image-caption pairs
+
+## Usage
+You can load this dataset using the Hugging Face Datasets library:
+
+```python
+from datasets import load_dataset
+dataset = load_dataset("theavenger/light-effect-dataset", split="train")
+```