gpt2

question_paper.py

@@ -1,10 +1,14 @@
-from transformers import pipeline
+from transformers import pipeline, set_seed
+generator = pipeline('text-generation', model='gpt2')
+set_seed(40)
+
 from fastapi import FastAPI
 from fastapi.middleware.cors import CORSMiddleware
 from fastapi.encoders import jsonable_encoder
 from fastapi.responses import JSONResponse
 
-pipe_flan = pipeline("text2text-generation", model="google/flan-t5-small", max_new_tokens=100)
+
+tokenizer.pad_token_id = tokenizer.eos_token_id
 
 app = FastAPI()
 
@@ -30,9 +34,9 @@ def llama(input):
 
     # tresponse = TOKENIZER.decode(tokens[0], skip_special_tokens=False)
     # print(tresponse)
-    output = pipe_flan(input)
+    outputs = generator(input, max_length=100, num_return_sequences=5)
     text = ""
-    for o in output:
+    for o in outputs:
         text = text + ' ' + o["generated_text"]
     response_message = {"message": text}
     json_response = jsonable_encoder(response_message)