Upload 2 files
Browse files- app.py +59 -0
- requirements.txt +0 -0
app.py
ADDED
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import torch
|
2 |
+
from transformers import AutoModelForSequenceClassification, AutoTokenizer
|
3 |
+
from azure.storage.blob import BlobServiceClient
|
4 |
+
from flask import Flask, request, jsonify
|
5 |
+
|
6 |
+
app = Flask(__name__)
|
7 |
+
|
8 |
+
# BERT model and tokenizer
|
9 |
+
model_name = "textattack/bert-base-uncased-yelp-polarity"
|
10 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
11 |
+
model = AutoModelForSequenceClassification.from_pretrained(model_name)
|
12 |
+
|
13 |
+
# Predict the category
|
14 |
+
def predict_category(input_text):
|
15 |
+
inputs = tokenizer(input_text, return_tensors="pt", padding=True, truncation=True)
|
16 |
+
with torch.no_grad():
|
17 |
+
logits = model(**inputs).logits
|
18 |
+
probabilities = logits.softmax(dim=1)
|
19 |
+
predicted_category = ["Documentation", "Content", "Memes"][torch.argmax(probabilities)]
|
20 |
+
return predicted_category
|
21 |
+
|
22 |
+
# Function to extract text from JSON and predict the category
|
23 |
+
def predict_category_from_json(json_data):
|
24 |
+
input_text = json_data.get('text', '')
|
25 |
+
category = predict_category(input_text)
|
26 |
+
return category
|
27 |
+
|
28 |
+
# Importing data from blob storage
|
29 |
+
def import_data_from_blob(blob_service_client, container_name, blob_name):
|
30 |
+
blob_client = blob_service_client.get_blob_client(container=container_name, blob=blob_name)
|
31 |
+
blob_data = blob_client.download_blob()
|
32 |
+
content = blob_data.readall()
|
33 |
+
return content
|
34 |
+
|
35 |
+
@app.route('/predict_category', methods=['POST'])
|
36 |
+
def predict_category_api():
|
37 |
+
try:
|
38 |
+
# Assuming JSON format with a key named 'text' that contains the text data.
|
39 |
+
json_data = request.get_json()
|
40 |
+
input_text = json_data.get('text', '')
|
41 |
+
|
42 |
+
# Predict the category
|
43 |
+
category = predict_category(input_text)
|
44 |
+
|
45 |
+
response = {'category': category}
|
46 |
+
return jsonify(response)
|
47 |
+
except Exception as e:
|
48 |
+
return jsonify({'error': str(e)})
|
49 |
+
|
50 |
+
if __name__ == '__main__':
|
51 |
+
# Azure Blob Storage connection string
|
52 |
+
connection_string = "DefaultEndpointsProtocol=https;AccountName=keywisestorage;AccountKey=uRzlCQwv/SSF6WgkEz0g83dBjnFrziSNNt8PIY5Nnt+OJic0v5xjPnO8ZMhb7SjyesYSOK79TbJ/+AStdLKiDw==;EndpointSuffix=core.windows.net"
|
53 |
+
blob_service_client = BlobServiceClient.from_connection_string(connection_string)
|
54 |
+
|
55 |
+
# Define your container and blob name
|
56 |
+
container_name = "keywisestorage"
|
57 |
+
blob_name = "pagescontainer"
|
58 |
+
|
59 |
+
app.run(host="0.0.0.0", port=5000)
|
requirements.txt
ADDED
Binary file (98 Bytes). View file
|
|