Lucas Tilford commited on
Commit
ff294e0
1 Parent(s): cdbcaf3

Add application file

Browse files
Files changed (1) hide show
  1. app.py +23 -0
app.py ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from transformers import AutoModelForQuestionAnswering, AutoTokenizer, pipeline
2
+ import gradio as gr
3
+ from datasets import load_dataset
4
+
5
+ # Load the UFO dataset from Hugging Face in chunks
6
+ dataset = load_dataset('your_dataset_name', split='train', streaming=True)
7
+
8
+ mdl_name = "deepset/roberta-base-squad2"
9
+ my_pipeline = pipeline('question-answering', model=mdl_name, tokenizer=mdl_name)
10
+
11
+ def answer_question(question):
12
+ # Iterate over chunks of the dataset
13
+ for chunk in dataset:
14
+ # Convert the chunk to a string to use as the context
15
+ context = ' '.join([str(item) for item in chunk])
16
+ response = my_pipeline({'question': question, 'context': context})
17
+ if response['score'] > 0.5: # Adjust this threshold as needed
18
+ return response
19
+
20
+ return "No answer found."
21
+
22
+ gr.Interface(answer_question, inputs="text", outputs="text").launch()
23
+