Rename chatv2.py to main.py

chatv2.py → main.py

@@ -1,5 +1,6 @@
 from flask import Flask, request, jsonify
 from huggingface_hub import InferenceClient
+import time
 
 # Initialize Flask app
 app = Flask(__name__)
@@ -47,11 +48,22 @@ def generate(prompt, history, temperature=0.9, max_new_tokens=256, top_p=0.95, r
     )
 
     output = ""
-    for token in response:
-        if hasattr(token, 'token') and hasattr(token.token, 'text'):
-            output += token.token.text
-        else:
-            print(f"Unexpected token structure: {token}", flush=True)
+    complete = False
+
+    # Use a loop to ensure response is fully received
+    while not complete:
+        for token in response:
+            if hasattr(token, 'token') and hasattr(token.token, 'text'):
+                output += token.token.text
+            else:
+                print(f"Unexpected token structure: {token}", flush=True)
+
+            # Check if the response seems complete
+            if token.token.text.endswith('</s>'):
+                complete = True
+                break
+        # Introduce a delay to handle streaming responses more smoothly
+        time.sleep(0.1)
 
     # Print AI response
     print(f"\nSema AI: {output}\n", flush=True)