Update main.py

main.py

@@ -18,6 +18,7 @@ parser.add_argument("--port", type=int, help="Set the port.(default: 7860)", def
 args = parser.parse_args()
 
 base_url = os.getenv('MODEL_BASE_URL')
+print(base_url)
 
 @app.route('/api/v1/models', methods=["GET", "POST"])
 @app.route('/v1/models', methods=["GET", "POST"])
@@ -44,9 +45,11 @@ def model_list():
 
 @app.route("/", methods=["GET"])
 def index():
-    return Response(f'QW2.5 OpenAI Compatible API<br><br>'+
-        f'Set "{os.getenv("SPACE_URL")}/api" as proxy (or API Domain) in your Chatbot.<br><br>'+
-        f'The complete API is: {os.getenv("SPACE_URL")}/api/v1/chat/completions')
+    print('index')
+    return Response(f"Hunyuan-Large OpenAI Compatible API<br><br>"+
+        f"Set '{os.getenv("SPACE_URL")}/api' as proxy (or API Domain) in your Chatbot.<br><br>"+
+        f"The complete API is: {os.getenv("SPACE_URL")}/api/v1/chat/completions<br><br>")
+        f"Don't set the Syetem Prompt. It will be ignored."
 
 @app.route("/api/v1/chat/completions", methods=["POST", "OPTIONS"])
 @app.route("/v1/chat/completions", methods=["POST", "OPTIONS"])
@@ -72,6 +75,7 @@ def chat_completions():
         message_size = len(messages)
 
         prompt = messages[-1].get("content")
+        
         for i in range(message_size - 1):
             role_this = messages[i].get("role")
             role_next = messages[i + 1].get("role")
@@ -89,22 +93,41 @@ def chat_completions():
         # print(f'{chat_history = }')
         # print(f'{prompt = }')
 
-        fn_index = 0
+        fn_index = 3
 
-        # gen a random char(11) hash
+        # gen a random char(10) hash
         chars = string.ascii_lowercase + string.digits
-        session_hash = "".join(random.choice(chars) for _ in range(11))
-
-        json_prompt = {
-            "data": [prompt, chat_history, system],
-            "fn_index": fn_index,
-            "session_hash": session_hash,
+        session_hash = "".join(random.choice(chars) for _ in range(10))
+        
+        single_prompt_data = {
+            'data': [
+                prompt,
+                [],
+            ],
+            'event_data': None,
+            'fn_index': 1,
+            'trigger_id': 5,
+            'session_hash': session_hash,
+        }
+        response = requests.post(f'{base_url}/gradio_api/run/predict', json=single_prompt_data)
+        
+        context_data = {
+            'data': [
+                None,
+                chat_history+[[prompt,None]]
+            ],
+            'event_data': None,
+            'fn_index': fn_index,
+            'trigger_id': 5,
+            'session_hash': session_hash,
         }
+        response = requests.post(f"{base_url}/gradio_api/queue/join", json=context_data)
+
+    def generate():        
 
-    def generate():
-        response = requests.post(f"{base_url}/queue/join", json=json_prompt)
-        url = f"{base_url}/queue/data?session_hash={session_hash}"
+        url = f"{base_url}/gradio_api/queue/data?session_hash={session_hash}"
         data = requests.get(url, stream=True)
+        #print(data.text)
 
         time_now = int(time.time())
 
@@ -143,7 +166,7 @@ def gen_res_data(data, time_now=0, start=False):
     if start:
         res_data["choices"][0]["delta"] = {"role": "assistant", "content": ""}
     else:
-        chat_pair = data["output"]["data"][1]
+        chat_pair = data["output"]["data"][0]
         if chat_pair == []:
             res_data["choices"][0]["finish_reason"] = "stop"
         else:
@@ -152,5 +175,5 @@ def gen_res_data(data, time_now=0, start=False):
 
 
 if __name__ == "__main__":
-    # app.run(host=args.host, port=args.port, debug=True)
-    gevent.pywsgi.WSGIServer((args.host, args.port), app).serve_forever()
+    #app.run(host=args.host, port=args.port, debug=True)
+    gevent.pywsgi.WSGIServer((args.host, args.port), app).serve_forever()