Noveramaaz
commited on
Commit
•
7be7efa
1
Parent(s):
9b94f9a
Update main.py
Browse files
main.py
CHANGED
@@ -1,3 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
# Function to lemmatize text
|
2 |
def lemmatize(text):
|
3 |
wordnet_lemmatizer = WordNetLemmatizer()
|
@@ -17,7 +45,7 @@ def preprocess_text(text):
|
|
17 |
text = lemmatize(text)
|
18 |
return text
|
19 |
|
20 |
-
# Load the model using FastAPI lifespan event so that
|
21 |
@asynccontextmanager
|
22 |
async def lifespan(app: FastAPI):
|
23 |
# Load the model from HuggingFace transformers library
|
@@ -28,14 +56,8 @@ async def lifespan(app: FastAPI):
|
|
28 |
# Clean up the model and release the resources
|
29 |
del sentiment_task
|
30 |
|
31 |
-
description = """
|
32 |
-
## Text Classification API
|
33 |
-
This app shows the sentiment of the text (positive, negative, or neutral).
|
34 |
-
Check out the docs for the `/analyze/{text}` endpoint below to try it out!
|
35 |
-
"""
|
36 |
-
|
37 |
# Initialize the FastAPI app
|
38 |
-
app = FastAPI(lifespan=lifespan
|
39 |
|
40 |
# Define the input data model
|
41 |
class TextInput(BaseModel):
|
@@ -46,40 +68,9 @@ class TextInput(BaseModel):
|
|
46 |
async def welcome():
|
47 |
return "Welcome to our Text Classification API"
|
48 |
|
49 |
-
# Redirect to the Swagger UI page
|
50 |
-
return RedirectResponse(url="/docs")
|
51 |
-
|
52 |
# Validate input text length
|
53 |
MAX_TEXT_LENGTH = 1000
|
54 |
|
55 |
-
# Define the sentiment analysis endpoint
|
56 |
-
@app.post('/analyze/{text}')
|
57 |
-
async def classify_text(text_input:TextInput):
|
58 |
-
try:
|
59 |
-
# Convert input data to JSON serializable dictionary
|
60 |
-
text_input_dict = jsonable_encoder(text_input)
|
61 |
-
# Validate input data using Pydantic model
|
62 |
-
text_data = TextInput(**text_input_dict) # Convert to Pydantic model
|
63 |
-
|
64 |
-
# Validate input text length
|
65 |
-
if len(text_input.text) > MAX_TEXT_LENGTH:
|
66 |
-
raise HTTPException(status_code=400, detail="Text length exceeds maximum allowed length")
|
67 |
-
elif len(text_input.text) == 0:
|
68 |
-
raise HTTPException(status_code=400, detail="Text cannot be empty")
|
69 |
-
except ValidationError as e:
|
70 |
-
# Handle validation error
|
71 |
-
raise HTTPException(status_code=422, detail=str(e))
|
72 |
-
|
73 |
-
try:
|
74 |
-
# Perform text classification
|
75 |
-
return sentiment_task(preprocess_text(text_input.text))
|
76 |
-
except ValueError as ve:
|
77 |
-
# Handle value error
|
78 |
-
raise HTTPException(status_code=400, detail=str(ve))
|
79 |
-
except Exception as e:
|
80 |
-
# Handle other server errors
|
81 |
-
raise HTTPException(status_code=500, detail=str(e))
|
82 |
-
|
83 |
# Define the sentiment analysis endpoint
|
84 |
@app.post('/analyze/{text}')
|
85 |
async def classify_text(text_input:TextInput):
|
@@ -129,8 +120,7 @@ class TextInput(BaseModel):
|
|
129 |
# Define the welcome endpoint
|
130 |
@app.get('/')
|
131 |
async def welcome():
|
132 |
-
|
133 |
-
return RedirectResponse(url="/docs")
|
134 |
|
135 |
# Validate input text length
|
136 |
MAX_TEXT_LENGTH = 1000
|
|
|
1 |
+
from contextlib import asynccontextmanager
|
2 |
+
from fastapi import FastAPI, HTTPException
|
3 |
+
from pydantic import BaseModel, ValidationError
|
4 |
+
from fastapi.encoders import jsonable_encoder
|
5 |
+
|
6 |
+
# TEXT PREPROCESSING
|
7 |
+
# --------------------------------------------------------------------
|
8 |
+
import re
|
9 |
+
import string
|
10 |
+
import nltk
|
11 |
+
nltk.download('punkt')
|
12 |
+
nltk.download('wordnet')
|
13 |
+
nltk.download('omw-1.4')
|
14 |
+
from nltk.stem import WordNetLemmatizer
|
15 |
+
|
16 |
+
# Function to remove URLs from text
|
17 |
+
def remove_urls(text):
|
18 |
+
return re.sub(r'http[s]?://\S+', '', text)
|
19 |
+
|
20 |
+
# Function to remove punctuations from text
|
21 |
+
def remove_punctuation(text):
|
22 |
+
regular_punct = string.punctuation
|
23 |
+
return str(re.sub(r'['+regular_punct+']', '', str(text)))
|
24 |
+
|
25 |
+
# Function to convert the text into lower case
|
26 |
+
def lower_case(text):
|
27 |
+
return text.lower()
|
28 |
+
|
29 |
# Function to lemmatize text
|
30 |
def lemmatize(text):
|
31 |
wordnet_lemmatizer = WordNetLemmatizer()
|
|
|
45 |
text = lemmatize(text)
|
46 |
return text
|
47 |
|
48 |
+
# Load the model using FastAPI lifespan event so that the model is loaded at the beginning for efficiency
|
49 |
@asynccontextmanager
|
50 |
async def lifespan(app: FastAPI):
|
51 |
# Load the model from HuggingFace transformers library
|
|
|
56 |
# Clean up the model and release the resources
|
57 |
del sentiment_task
|
58 |
|
|
|
|
|
|
|
|
|
|
|
|
|
59 |
# Initialize the FastAPI app
|
60 |
+
app = FastAPI(lifespan=lifespan)
|
61 |
|
62 |
# Define the input data model
|
63 |
class TextInput(BaseModel):
|
|
|
68 |
async def welcome():
|
69 |
return "Welcome to our Text Classification API"
|
70 |
|
|
|
|
|
|
|
71 |
# Validate input text length
|
72 |
MAX_TEXT_LENGTH = 1000
|
73 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
74 |
# Define the sentiment analysis endpoint
|
75 |
@app.post('/analyze/{text}')
|
76 |
async def classify_text(text_input:TextInput):
|
|
|
120 |
# Define the welcome endpoint
|
121 |
@app.get('/')
|
122 |
async def welcome():
|
123 |
+
return "Welcome to our Text Classification API"
|
|
|
124 |
|
125 |
# Validate input text length
|
126 |
MAX_TEXT_LENGTH = 1000
|