change model

app.py

@@ -1,7 +1,7 @@
 from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM
 from transformers.utils import logging
 import gradio as gr
-#import spaces
+import spaces
 
 # Define the logger instance for the transformers library
 logger = logging.get_logger("transformers")
@@ -14,23 +14,23 @@ model = AutoModelForCausalLM.from_pretrained(model_name,device_map="auto",trust_
 
 #transfer model on GPU
 #model.to("cuda")
-pipe = pipeline("text-generation", model=model_name, tokenizer=tokenizer, 
-        max_new_tokens=512,
-        do_sample=True,
-        temperature=0.7,
-        top_p=0.95,
-        top_k=40,
-        repetition_penalty=1.1)
+# pipe = pipeline("text-generation", model=model_name, tokenizer=tokenizer, 
+#         max_new_tokens=512,
+#         do_sample=True,
+#         temperature=0.7,
+#         top_p=0.95,
+#         top_k=40,
+#         repetition_penalty=1.1)
 
 # Generate text using the model and tokenizer
-#@spaces.GPU(duration=60)
+@spaces.GPU(duration=60)
 def generate_text(input_text):
-    #input_ids = tokenizer.encode(input_text, return_tensors="pt")#.to("cuda")
+    input_ids = tokenizer.encode(input_text, return_tensors="pt")#.to("cuda")
     #attention_mask = input_ids.ne(tokenizer.pad_token_id).long()
-    #output = model.generate(input_ids, max_new_tokens=512, top_k=50, top_p=0.95, temperature=0.7, do_sample=True)# attention_mask=attention_mask, max_length=100, num_return_sequences=1, no_repeat_ngram_size=2, top_k=50, top_p=0.95, temperature=0.7, do_sample=True)
+    output = model.generate(input_ids, max_new_tokens=512, top_k=50, top_p=0.95, temperature=0.7, do_sample=True)# attention_mask=attention_mask, max_length=100, num_return_sequences=1, no_repeat_ngram_size=2, top_k=50, top_p=0.95, temperature=0.7, do_sample=True)
     #output = model.generate(input_ids) #, attention_mask=attention_mask, max_length=100, num_return_sequences=1, no_repeat_ngram_size=2, top_k=50, top_p=0.95, temperature=0.7, do_sample=True)
-    #return tokenizer.decode(output[0])
-    return pipe(input_text)[0]["generated_text"]
+    return tokenizer.decode(output[0])
+    #return pipe(input_text)[0]["generated_text"]
 
 interface = gr.Interface(fn=generate_text, inputs="text", outputs="text",title="TeLLMyStory",description="Enter your story idea and the model will generate the story based on it.")
 interface.launch()