README: load recaptions with the parquet builder

README.md

@@ -10,10 +10,6 @@ configs:
   data_files:
   - split: train
     path: "*.tar"
-- config_name: recaptions
-  data_files:
-  - split: train
-    path: "recaptions/*.parquet"
 ---
 
 # BLIP3o-Recap
@@ -77,7 +73,11 @@ ds = load_dataset("webdataset", data_files=glob.glob("/path/to/shards/*.tar"), s
 ```python
 from datasets import load_dataset
 
-recaps = load_dataset("IntelligenceLab/BLIP3o-Recap", "recaptions", split="train")
+recaps = load_dataset(
+    "parquet",
+    data_files="hf://datasets/IntelligenceLab/BLIP3o-Recap/recaptions/*.parquet",
+    split="train",
+)
 ```
 
 ### Generation