updates

models/base.py

@@ -1,6 +1,6 @@
 import os
 import requests
-
+import gradio as gr
 from conversation import Conversation
 
 
@@ -38,5 +38,12 @@ class BaseModel:
             parameters.update(custom_generation_params)
         payload = {'instances': [text], "parameters": parameters}
         resp = requests.post(api, json=payload, timeout=600)
-        assert resp.status_code == 200, (resp.content, resp.status_code)
+        if resp.status_code != 200:
+            gr.Error(f"Endpoint returned code: {resp.status_code}.\n"
+                     f"Content: {resp.content}.\n"
+                     f"Notes:\n"
+                     f"- we have scale-to-zero enabled, so please wait for some minutes and try again.\n"
+                     f"- probably the response generated by the model is to big, try changing max_new_tokens.\n"
+                     f"- if nothing helps — report the problem.")
+        # assert resp.status_code == 200, (resp.content, resp.status_code)
         return resp.json()["predictions"][0].strip()