Spaces:
Sleeping
Sleeping
Update load_model.py
Browse files- load_model.py +4 -34
load_model.py
CHANGED
@@ -1,39 +1,9 @@
|
|
1 |
import os
|
|
|
2 |
from dotenv import load_dotenv
|
3 |
-
from transformers import TFBertForSequenceClassification, BertTokenizerFast
|
4 |
|
5 |
-
# Load environment variables from .env file
|
6 |
load_dotenv()
|
|
|
7 |
|
8 |
-
def load_model(
|
9 |
-
|
10 |
-
# Load TensorFlow model from Hugging Face
|
11 |
-
model = TFBertForSequenceClassification.from_pretrained(model_name, use_auth_token=os.getenv('API_KEY'), from_tf=True)
|
12 |
-
except OSError:
|
13 |
-
raise ValueError("Model loading failed.")
|
14 |
-
return model
|
15 |
-
|
16 |
-
def load_tokenizer(model_name):
|
17 |
-
tokenizer = BertTokenizerFast.from_pretrained(model_name, use_auth_token=os.getenv('API_KEY'))
|
18 |
-
return tokenizer
|
19 |
-
|
20 |
-
def predict(text, model, tokenizer):
|
21 |
-
inputs = tokenizer(text, return_tensors="tf")
|
22 |
-
outputs = model(**inputs)
|
23 |
-
return outputs
|
24 |
-
|
25 |
-
def main():
|
26 |
-
model_name = os.getenv('MODEL_PATH')
|
27 |
-
if model_name is None:
|
28 |
-
raise ValueError("MODEL_PATH environment variable not set or is None")
|
29 |
-
|
30 |
-
model = load_model(model_name)
|
31 |
-
tokenizer = load_tokenizer(model_name)
|
32 |
-
|
33 |
-
# Example prediction
|
34 |
-
text = "Sample input text"
|
35 |
-
result = predict(text, model, tokenizer)
|
36 |
-
print(result)
|
37 |
-
|
38 |
-
if __name__ == "__main__":
|
39 |
-
main()
|
|
|
1 |
import os
|
2 |
+
import tensorflow as tf
|
3 |
from dotenv import load_dotenv
|
|
|
4 |
|
|
|
5 |
load_dotenv()
|
6 |
+
model_path = os.getenv('MODEL_PATH')
|
7 |
|
8 |
+
def load_model():
|
9 |
+
return tf.keras.models.load_model(model_path)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|