Update app.py

app.py

@@ -55,7 +55,7 @@ def summarize(inp, model=model, tokenizer=tokenizer, device=device):
         inputs = tokenizer(inp, return_tensors="pt", padding=True).to(device)
         batch_size, sequence_length = inputs["input_ids"].shape
         max_cache_length = 1024
-        max_new_tokens = 20
+        max_new_tokens = 64
         # setup static cache
         past_key_values = StaticCache(
             config=model.config,
@@ -105,5 +105,5 @@ def summarize(inp, model=model, tokenizer=tokenizer, device=device):
       
         return response
 
-gr.Interface(fn=summarize, inputs=gr.Textbox(lines=7, label="Input Text"), outputs="text", title="gemma-2b Demo").launch(inline=False)
+gr.Interface(fn=summarize, inputs=gr.Textbox(lines=7, label="Input Text"), outputs="text", title="gemma-2b simple TPU demo").launch(inline=False)