Update app.py

app.py

@@ -60,9 +60,10 @@ def query(payload):
 
 
 def inference(input_sentence, max_length, sample_or_greedy, raw_text=True):
+    do_sample = sample_or_greedy == "sampling"
     payload = {
         "inputs": input_sentence,
-        "do_sample": True,
+        "do_sample": do_sample,
         # "max_new_tokens": max_length
     }
 
@@ -113,7 +114,7 @@ gr.Interface(
     [
         gr.inputs.Textbox(label="Input"),
         gr.inputs.Radio([64], default=64, label="Tokens to generate"),
-        gr.inputs.Radio(["sampling"], label="Sample or greedy", default="sampling"),
+        gr.inputs.Radio(["sampling", "greedy"], label="Sample or greedy", default="sampling"),
         gr.Checkbox(label="Just output raw text", value=True),
     ],
     ["image", "text"],