Update README.md

README.md

@@ -152,7 +152,7 @@ _Main split for Hugging Face: JSONL format (see above for statistics)._
   "sci_fields": {},
   "incl_fields": {}
 }
-
+```
 
 
 ===================================================================================================================
@@ -163,9 +163,8 @@ The Hugging Face infrastructure will automatically use the efficient Parquet fil
 ### Install dependencies (if needed):
 
 ```bash
-pip install datasets```
-
-
+pip install datasets
+```
 
 ## Load the full dataset (Parquet, recommended)
 
@@ -175,9 +174,8 @@ pip install datasets```
 ds = load_dataset("SkyWhal3/ClinVar-STXBP1-NLP-Dataset")
 
 # Access examples
-print(ds["train"][0])```
-
-
+print(ds["train"][0])
+```
 
 ## To force JSONL loading (if you prefer the original format):
 
@@ -189,8 +187,8 @@ ds = load_dataset(
     data_files="ClinVar-STXBP1-NLP-Dataset.jsonl",
     split="train"
 )
-print(ds[0])```
-
+print(ds[0])
+```
 
 ## Other ways to use the data
    Load all Parquet shards with pandas
@@ -202,29 +200,29 @@ import glob
 parquet_files = glob.glob("default/train/*.parquet")
 df = pd.concat([pd.read_parquet(pq) for pq in parquet_files], ignore_index=True)
 print(df.shape)
-print(df.head())```
-
+print(df.head())
+```
 
 ## Filter for a gene (e.g., STXBP1)
 
 ```df = pd.read_parquet("default/train/0000.parquet")
 stxbp1_df = df[df["gene"] == "STXBP1"]
-print(stxbp1_df.head())```
-
+print(stxbp1_df.head())
+```
 
 ## Randomly sample a subset
 
 ```sample = df.sample(n=5, random_state=42)
-print(sample)```
-
+print(sample)
+```
 
 ## Load with Polars (for high performance)
 
 ```import polars as pl
 
 df = pl.read_parquet("default/train/0000.parquet")
-print(df.head())```
-
+print(df.head())
+```
 
 ## Query with DuckDB (SQL-style)
 
@@ -232,12 +230,12 @@ print(df.head())```
 
 con = duckdb.connect()
 df = con.execute("SELECT * FROM 'default/train/0000.parquet' WHERE gene='STXBP1' LIMIT 5").df()
-print(df)```
-
-
-```## Streaming mode with 🤗 Datasets
+print(df)
+```
 
+## Streaming mode with 🤗 Datasets
 
 ds = load_dataset("SkyWhal3/ClinVar-STXBP1-NLP-Dataset", split="train", streaming=True)
 for record in ds.take(5):
-    print(record)```
+    print(record)
+```