Update README.md

README.md

@@ -47,10 +47,9 @@ English captions were translated into Hungarian using a large-scale automated tr
 ## Dataset Structure
 
 Each sample contains:
-- `image`: RGB image (JPEG)
-- `caption_en`: English caption
-- `caption_hu`: Hungarian caption
-- `source`: Original dataset identifier
+- `img`: RGB image (JPEG)
+- `en_cap`: English caption
+- `hu_cap`: Hungarian caption
 
 Data is stored in **Parquet format**, sharded into ~1M-sample files for efficient streaming.
 
@@ -58,12 +57,11 @@ Data is stored in **Parquet format**, sharded into ~1M-sample files for efficien
 
 ## Dataset Creation Pipeline
 
-1. Caption normalization and cleaning
-2. Automated English → Hungarian translation
-3. Parallel image downloading
-4. Image validation and filtering
-5. Post-processing and shard merging
-6. Upload to Hugging Face Hub
+1. Automated English → Hungarian translation
+2. Parallel image downloading
+3. Image validation and filtering
+4. Post-processing and shard merging
+5. Upload to Hugging Face Hub
 
 The pipeline is fully automated and distributed across parallel workers.