add install.sh

app.py

@@ -76,11 +76,15 @@ def respond(
 
     model_inputs = tokenizer([text], return_tensors="pt", return_token_type_ids=False).to(model.device)
 
-    model_inputs.update(dict(max_new_tokens=max_tokens,
-                             temperature = temperature,
-                             top_p = top_p,
+    print(f"max_new_tokens={max_tokens}, temperature={temperature}, top_p={top_p}")
+    model_inputs.update(
+        dict(max_new_tokens=max_tokens,
+                             streamer = streamer,
+                             temperature = float(temperature),
+                             top_p = float(top_p),
                              presence_penalty = 1.5,
-                             streamer=streamer))
+             )
+    )
 
     # Start a separate thread for model generation to allow streaming output
     thread = Thread(