new

app.py

@@ -28,6 +28,8 @@ def query(payload):
 
 
 def inference(input_sentence, max_length, sample_or_greedy, seed=42):
+
+    print("max_length", max_length)
     if sample_or_greedy == "Sample":
         parameters = {
             "max_new_tokens": max_length,
@@ -52,6 +54,7 @@ def inference(input_sentence, max_length, sample_or_greedy, seed=42):
     model_name = 'bigscience/bloomz-560m'
     pipe = pipeline("text-generation", model = model_name, tokenizer = model_name)
 
+    '''
     if sample_or_greedy == "Sample":
         res = pipe(input_sentence,  
             max_new_tokens = max_length, 
@@ -69,8 +72,10 @@ def inference(input_sentence, max_length, sample_or_greedy, seed=42):
             early_stopping = False, 
             eos_token_id = None
         )
+    '''
+
+    res = pipe(payload['inputs'], **payload['parameters'])
 
-    #res = pipe(payload['inputs'], **payload['parameters'])
     #data = query(payload)
     #if "error" in data:
     #    return (None, None, f"<span style='color:red'>ERROR: {data['error']} </span>")