Update app.py

app.py

@@ -1,6 +1,8 @@
 #import libraries and dependencies
-import gradio as gr
 #from gradio.mix import Parallel
+
+import gradio as gr
+import torch
 from transformers import pipeline
 
 #instantiate variables as strings
@@ -18,7 +20,9 @@ from transformers import pipeline
 #]
 
 #instantiate variables as functions
-pipe = pipeline("text-generation", model='EleutherAI/gpt-neo-2.7B', trust_remote_code=True)
+#pipe = pipeline("text-generation", model='EleutherAI/gpt-neo-2.7B', trust_remote_code=True)
+
+ans = pipeline(model="databricks/dolly-v2-3b", torch_dtype=torch.bfloat16, trust_remote_code=True, device_map="auto")
 
 #model1 = gr.Interface.load("huggingface/bigscience/bloom-560m")
 #model2 = gr.Interface.load("huggingface/google/flan-t5-xl")
@@ -27,7 +31,13 @@ pipe = pipeline("text-generation", model='EleutherAI/gpt-neo-2.7B', trust_remote
 
 #togethercomputer/GPT-NeoXT-Chat-Base-20B
 #decapoda-research/llama-7b-hf
+
 #define functions
+
+def answer(query):
+  out=ans(query)
+  return out
+
 #def complete_with_gpt(text):
 #    # Use the last 50 characters of the text as context
 #    return text[:-50] + model4(text[-50:])
@@ -48,5 +58,7 @@ pipe = pipeline("text-generation", model='EleutherAI/gpt-neo-2.7B', trust_remote
 
 #if __name__ == "__main__":
 #    demo1.launch(debug=True)
+#gr.Interface.from_pipeline(pipe).launch()
 
-gr.Interface.from_pipeline(pipe).launch()
+Demo = gr.Interface(fn=answer,inputs='text',outputs='text',examples=[['What is the capital of India ?']])
+Demo.launch()