Modify app.py with non-error typehints

app.py

@@ -24,24 +24,38 @@ sentiment_model = AutoModelForSequenceClassification.from_pretrained(sentiment_m
 sentiment_model.eval()
 
 
+# ==============================
+# FUNCTION (SIGNATURE UNCHANGED)
+# Input: DataFrame
+# Output: DataFrame
+# ==============================
+
 def nlp_encode_sentence(df: pd.DataFrame) -> pd.DataFrame:
 
+    # Convert safely in case input is list-of-lists
+    df = pd.DataFrame(df, columns=["value"])
+
     feature_rows = []
 
     for sentence in df["value"]:
 
-        inputs = tokenizer(sentence, return_tensors="pt", truncation=True, padding=True)
+        inputs = tokenizer(
+            sentence,
+            return_tensors="pt",
+            truncation=True,
+            padding=True
+        )
 
         with torch.no_grad():
             outputs = bert_model(**inputs)
 
         cls_embedding = outputs.last_hidden_state[:, 0, :].squeeze().numpy()
 
-        embedding_mean = np.mean(cls_embedding)
-        embedding_median = np.median(cls_embedding)
-        embedding_std = np.std(cls_embedding)
-        embedding_min = np.min(cls_embedding)
-        embedding_max = np.max(cls_embedding)
+        embedding_mean = float(np.mean(cls_embedding))
+        embedding_median = float(np.median(cls_embedding))
+        embedding_std = float(np.std(cls_embedding))
+        embedding_min = float(np.min(cls_embedding))
+        embedding_max = float(np.max(cls_embedding))
 
         sentiment_inputs = sentiment_tokenizer(
             sentence,
@@ -54,27 +68,31 @@ def nlp_encode_sentence(df: pd.DataFrame) -> pd.DataFrame:
             sentiment_outputs = sentiment_model(**sentiment_inputs)
 
         probs = softmax(sentiment_outputs.logits.numpy()[0])
-        sentiment_score = probs[2] - probs[0]
+        sentiment_score = float(probs[2] - probs[0])
 
-        feature_rows.append({
-            "embedding_mean": embedding_mean,
-            "embedding_median": embedding_median,
-            "embedding_std": embedding_std,
-            "embedding_min": embedding_min,
-            "embedding_max": embedding_max,
-            "sentiment_score": sentiment_score
-        })
+        feature_rows.append([
+            embedding_mean,
+            embedding_median,
+            embedding_std,
+            embedding_min,
+            embedding_max,
+            sentiment_score
+        ])
 
-    features_df = pd.DataFrame(feature_rows)
+    # Return as array (important for Gradio compatibility)
+    return feature_rows
 
-    return features_df
 
+# ==============================
+# GRADIO INTERFACE
+# ==============================
 
 demo = gr.Interface(
     fn=nlp_encode_sentence,
-    inputs=gr.Dataframe(),
-    outputs=gr.Dataframe(),
+    inputs=gr.Dataframe(type="array"),
+    outputs=gr.Dataframe(type="array"),
     api_name="encode"
 )
 
-demo.launch()
+if __name__ == "__main__":
+    demo.launch()