Update README.md

README.md

@@ -318,6 +318,21 @@ urlretrieve(table_url, 'metadata.parquet')
 # Read the table using Pandas
 metadata_df = pd.read_parquet('metadata.parquet')
 ```
+### Method 4.
+
+This method works fine:
+```python
+from huggingface_hub import hf_hub_download
+import pandas as pd
+
+REPO_ID = "poloclub/diffusiondb" # dataset repo
+FILENAME = "images/part-000001.zip" #first subset of dataset
+
+hf_hub_download(repo_id=REPO_ID, filename=FILENAME, repo_type="dataset", local_dir=local_path)
+
+# then you will need the json file which will be located in local_path/part-000001.json
+metadata = pd.read_json("local_path/part-000001.json")
+```
 
 ## Dataset Creation