feat: set casting to str by default

app.py

@@ -1,3 +1,5 @@
+import time
+
 import duckdb
 import gradio as gr
 import polars as pl
@@ -56,18 +58,24 @@ def vectorize_dataset(split: str, column: str):
     global df
     global ds
     df = ds[split].to_polars()
-    embeddings = model.encode(df[column], max_length=512)
-    df = df.with_columns(pl.Series(embeddings).alias("embeddings"))
+    embeddings = model.encode(df[column].cast(str), max_length=512)
+    df = df.with_columns(pl.Series(embeddings).alias(f"{column}_embeddings"))
 
 
-def run_query(query: str):
+def run_query(query: str, column: str):
     global df
+    while f"{column}_embeddings" not in df.columns:
+        sleeper = 5
+        gr.Info(
+            f"Waiting for vectorization to complete... ({sleeper}s)", duration=sleeper
+        )
+        time.sleep(sleeper)
     vector = model.encode(query)
     df_results = duckdb.sql(
         query=f"""
         SELECT *
         FROM df
-        ORDER BY array_cosine_distance(embeddings, {vector.tolist()}::FLOAT[256])
+        ORDER BY array_cosine_distance({column}_embeddings, {vector.tolist()}::FLOAT[256])
         LIMIT 5
         """
     ).to_df()
@@ -152,6 +160,8 @@ with gr.Blocks() as demo:
         fn=show_components, outputs=[query_input, btn_run]
     )
 
-    btn_run.click(fn=run_query, inputs=query_input, outputs=results_output)
+    btn_run.click(
+        fn=run_query, inputs=[query_input, column_dropdown], outputs=results_output
+    )
 
 demo.launch()