s

app.py

@@ -25,7 +25,9 @@ device = ("cuda"
 logging.info(C("[INFO] "f"device = {device}"))
 
 # ------------------ INITITALIZE ------------------- #
-@st.cache
+@st.cache(
+        suppress_st_warning=True
+)
 def model_init():
     
     
@@ -68,8 +70,10 @@ try:
             [texts_out] = model.generate(
                 **tokenizer(
                     prompt, return_tensors="pt",
-                    max_new_tokens=200,
-                ).to(device))
+                    
+                ).to(device),
+                max_new_tokens=200,
+                )
         output_text = tokenizer.decode(texts_out)
         st.balloons()
         st.markdown(output_text)