Upload app.py
Browse files
app.py
ADDED
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
import pickle
|
3 |
+
import tensorflow as tf
|
4 |
+
import gradio as gr
|
5 |
+
import pandas as pd
|
6 |
+
from sklearn.preprocessing import StandardScaler
|
7 |
+
|
8 |
+
data_heading = ['longitude', 'latitude', 'housing_median_age', 'total_rooms',
|
9 |
+
'total_bedrooms', 'population', 'households', 'median_income',
|
10 |
+
'median_house_value']
|
11 |
+
|
12 |
+
# Model and scaler loading
|
13 |
+
with open("./model/scaler_sklearn.pkl", "rb") as f:
|
14 |
+
scaler = pickle.load(f)
|
15 |
+
loaded_model = tf.keras.saving.load_model('./model/house_value_model.keras')
|
16 |
+
|
17 |
+
def test_ml_model(longitude, latitude, housing_median_age, total_rooms, total_bedrooms, population, households, median_income, median_house_value):
|
18 |
+
|
19 |
+
df_test = pd.DataFrame(data=[longitude, latitude, housing_median_age,
|
20 |
+
total_rooms, total_bedrooms, population,
|
21 |
+
households, median_income, median_house_value], columns=data_heading)
|
22 |
+
df_test_norm = pd.DataFrame(scaler(df_test), columns=data_heading)
|
23 |
+
|
24 |
+
result = loaded_model.predict(df_test_norm)
|
25 |
+
return (f'predicted: {result}')
|
26 |
+
|
27 |
+
demo = gr.Interface(fn=test_ml_model,
|
28 |
+
inputs=[gr.Number(value=0.0), gr.Number(value=0.0), gr.Number(value=0.0),
|
29 |
+
gr.Number(value=0.0), gr.Number(value=0.0), gr.Number(value=0.0),
|
30 |
+
gr.Number(value=0.0), gr.Number(value=0.0), gr.Number(value=0.0),
|
31 |
+
gr.Number(value=0.0),],
|
32 |
+
outputs="text",
|
33 |
+
description="A sample linear regressor solution.",
|
34 |
+
title='Synthetic Data Linear Regressor Solution')
|
35 |
+
|
36 |
+
demo.launch()
|