Spaces:
Sleeping
Sleeping
Update tasks/text.py
Browse files- tasks/text.py +8 -8
tasks/text.py
CHANGED
@@ -2,7 +2,7 @@ from fastapi import APIRouter
|
|
2 |
from datetime import datetime
|
3 |
from datasets import load_dataset
|
4 |
from sklearn.feature_extraction.text import TfidfVectorizer
|
5 |
-
from sklearn.
|
6 |
from sklearn.metrics import accuracy_score
|
7 |
|
8 |
from .utils.evaluation import TextEvaluationRequest
|
@@ -10,18 +10,18 @@ from .utils.emissions import tracker, clean_emissions_data, get_space_info
|
|
10 |
|
11 |
router = APIRouter()
|
12 |
|
13 |
-
DESCRIPTION = "
|
14 |
-
ROUTE = "/
|
15 |
|
16 |
@router.post(ROUTE, tags=["Text Task"],
|
17 |
description=DESCRIPTION)
|
18 |
-
async def
|
19 |
"""
|
20 |
Evaluate text classification for climate disinformation detection.
|
21 |
|
22 |
-
Current Model:
|
23 |
- Uses TF-IDF for text vectorization
|
24 |
-
- Trains and evaluates a
|
25 |
"""
|
26 |
# Get space info
|
27 |
username, space_url = get_space_info()
|
@@ -60,8 +60,8 @@ async def evaluate_text(request: TextEvaluationRequest):
|
|
60 |
train_vectors = vectorizer.fit_transform(train_texts)
|
61 |
test_vectors = vectorizer.transform(test_texts)
|
62 |
|
63 |
-
# Train
|
64 |
-
model =
|
65 |
model.fit(train_vectors, train_labels)
|
66 |
|
67 |
# Start tracking emissions
|
|
|
2 |
from datetime import datetime
|
3 |
from datasets import load_dataset
|
4 |
from sklearn.feature_extraction.text import TfidfVectorizer
|
5 |
+
from sklearn.svm import SVC
|
6 |
from sklearn.metrics import accuracy_score
|
7 |
|
8 |
from .utils.evaluation import TextEvaluationRequest
|
|
|
10 |
|
11 |
router = APIRouter()
|
12 |
|
13 |
+
DESCRIPTION = "SVM Text Classifier with TF-IDF"
|
14 |
+
ROUTE = "/text_svm"
|
15 |
|
16 |
@router.post(ROUTE, tags=["Text Task"],
|
17 |
description=DESCRIPTION)
|
18 |
+
async def evaluate_text_svm(request: TextEvaluationRequest):
|
19 |
"""
|
20 |
Evaluate text classification for climate disinformation detection.
|
21 |
|
22 |
+
Current Model: SVM Classifier
|
23 |
- Uses TF-IDF for text vectorization
|
24 |
+
- Trains and evaluates a Support Vector Machine (SVM) model
|
25 |
"""
|
26 |
# Get space info
|
27 |
username, space_url = get_space_info()
|
|
|
60 |
train_vectors = vectorizer.fit_transform(train_texts)
|
61 |
test_vectors = vectorizer.transform(test_texts)
|
62 |
|
63 |
+
# Train SVM Classifier
|
64 |
+
model = SVC(kernel="linear", probability=True)
|
65 |
model.fit(train_vectors, train_labels)
|
66 |
|
67 |
# Start tracking emissions
|