|
|
|
|
|
|
|
|
|
import gradio as gr |
|
from transformers import pipeline |
|
|
|
"""# Loading Model Name""" |
|
|
|
model_name = "deepset/roberta-base-squad2" |
|
|
|
"""# Get Predictions |
|
|
|
""" |
|
|
|
nlu = pipeline('question-answering', model=model_name, tokenizer=model_name) |
|
|
|
def func(context, question): |
|
input = { |
|
'question':question, |
|
'context':context |
|
} |
|
res = nlu(input) |
|
return res["answer"] |
|
|
|
descr = "This is a question and Answer Web app, you give it a context and ask it questions based on the context provided" |
|
|
|
app = gr.Interface(fn=func, inputs=[gr.inputs.Textbox(lines=3, placeholder="put in your context here..."),"text"], outputs="text", title="Question Answer App", description=descr) |
|
|
|
app.launch() |
