Spaces:
Sleeping
Sleeping
Commit
·
f5ecfb4
1
Parent(s):
eb7dd0e
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,61 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from transformers import AutoTokenizer, AutoModelForTokenClassification
|
2 |
+
from transformers import pipeline
|
3 |
+
import gradio as gr
|
4 |
+
|
5 |
+
tokenizer = AutoTokenizer.from_pretrained("valurank/bert-base-NER")
|
6 |
+
model = AutoModelForTokenClassification.from_pretrained("valurank/bert-base-NER")
|
7 |
+
nlp = pipeline("ner", model=model, tokenizer=tokenizer, aggregation_strategy="simple")
|
8 |
+
|
9 |
+
def rename_group(output_list):
|
10 |
+
final_output = []
|
11 |
+
|
12 |
+
for output in output_list:
|
13 |
+
output["entity"] = output["entity_group"]
|
14 |
+
del output["entity_group"]
|
15 |
+
|
16 |
+
final_output.append(output)
|
17 |
+
|
18 |
+
return final_output
|
19 |
+
|
20 |
+
def remove_prefix(word, prefix):
|
21 |
+
if prefix in word:
|
22 |
+
return word.split(prefix, 1)[1]
|
23 |
+
return " " + word
|
24 |
+
|
25 |
+
def join_results(results):
|
26 |
+
joined_results = []
|
27 |
+
|
28 |
+
for result in results:
|
29 |
+
if "##" in result["word"] and joined_results:
|
30 |
+
joined_results[-1]["end"] = result["end"]
|
31 |
+
joined_results[-1]["word"] += remove_prefix(result["word"], "##")
|
32 |
+
joined_results[-1]["score"] = min(joined_results[-1]["score"], result["score"])
|
33 |
+
else:
|
34 |
+
joined_results.append(result)
|
35 |
+
|
36 |
+
return joined_results
|
37 |
+
|
38 |
+
examples = [
|
39 |
+
""" Texas A&M professor used chatbot chatbot to assess students' grades.
|
40 |
+
The OpenAI chatbot is actually called ChatGPT and claims to have written every paper written by the bot.
|
41 |
+
The bot isn’t made to detect material composed by AI, or even material produced by itself.
|
42 |
+
Texas A&M University-Commerce said they are investigating the incident and developing policies related to AI in the classroom.
|
43 |
+
The university denied that anyone had received a failing grade.
|
44 |
+
The school also confirmed that several students had been cleared of any academic dishonesty.
|
45 |
+
The use of AI in coursework is a rapidly changing issue that confronts all learning institutions."""
|
46 |
+
]
|
47 |
+
|
48 |
+
def ner(text):
|
49 |
+
output = nlp(text)
|
50 |
+
output = join_results(output)
|
51 |
+
output = rename_group(output)
|
52 |
+
|
53 |
+
return {"text": text, "entities": output}
|
54 |
+
|
55 |
+
demo = gr.Interface(ner,
|
56 |
+
gr.Textbox(placeholder="Enter sentence here..."),
|
57 |
+
gr.HighlightedText(),
|
58 |
+
examples=examples)
|
59 |
+
|
60 |
+
if __name__ == '__main__':
|
61 |
+
demo.launch(debug=True)
|