Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
@@ -1,7 +1,81 @@
|
|
1 |
import gradio as gr
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
|
3 |
-
|
4 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
5 |
|
6 |
-
|
7 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
import gradio as gr
|
2 |
+
import tensorflow
|
3 |
+
import numpy as np
|
4 |
+
import cv2 as cv
|
5 |
+
import requests
|
6 |
+
import time
|
7 |
+
import os
|
8 |
|
9 |
+
host = os.environ.get("host")
|
10 |
+
code = os.environ.get("code")
|
11 |
+
model = os.environ.get("model")
|
12 |
+
data = None
|
13 |
+
model = None
|
14 |
+
image = None
|
15 |
+
prediction = None
|
16 |
+
labels = None
|
17 |
+
max_label_index = None
|
18 |
+
max_prediction_value = -1
|
19 |
|
20 |
+
print('START')
|
21 |
+
np.set_printoptions(suppress=True)
|
22 |
+
|
23 |
+
model = tensorflow.keras.models.load_model('keras_model.h5')
|
24 |
+
data = np.ndarray(shape=(1, 224, 224, 3), dtype=np.float32)
|
25 |
+
|
26 |
+
with open("labels.txt", "r") as file:
|
27 |
+
labels = file.read().splitlines()
|
28 |
+
|
29 |
+
def classify(image_path):
|
30 |
+
try:
|
31 |
+
image = cv.imread(image_path)
|
32 |
+
image = cv.resize(image, (224, 224))
|
33 |
+
image_array = np.asarray(image)
|
34 |
+
normalized_image_array = (image_array.astype(np.float32) / 127.0) - 1
|
35 |
+
data[0] = normalized_image_array
|
36 |
+
prediction = model.predict(data)
|
37 |
+
|
38 |
+
print('Prediction')
|
39 |
+
|
40 |
+
for i, label in enumerate(labels):
|
41 |
+
prediction_value = float(prediction[0][i])
|
42 |
+
rounded_value = round(prediction_value, 2)
|
43 |
+
print(f'{label}: {rounded_value}')
|
44 |
+
|
45 |
+
if prediction_value > max_prediction_value:
|
46 |
+
max_label_index = i
|
47 |
+
max_prediction_value = prediction_value
|
48 |
+
|
49 |
+
if max_label_index is not None:
|
50 |
+
max_label = labels[max_label_index].split(' ', 1)[1]
|
51 |
+
print(f'Maximum Prediction: {max_label} with a value of {round(max_prediction_value, 2)}')
|
52 |
+
|
53 |
+
time.sleep(1)
|
54 |
+
print("\nWays to dispose this waste: " + max_label)
|
55 |
+
payload = [
|
56 |
+
{"role": "system", "content": "You are a helpful assistant."},
|
57 |
+
{"role": "user", "content": "Give me the steps to dispose this waste in bulleting points 5 max: " + "Plastic"}
|
58 |
+
]
|
59 |
+
|
60 |
+
response = requests.post(host, json={
|
61 |
+
"messages": payload,
|
62 |
+
"model": model,
|
63 |
+
"temperature": 0.5,
|
64 |
+
"presence_penalty": 0,
|
65 |
+
"frequency_penalty": 0,
|
66 |
+
"top_p": 1
|
67 |
+
}).json()
|
68 |
+
|
69 |
+
return response["choices"][0]["message"]["content"]
|
70 |
+
|
71 |
+
except Exception as e:
|
72 |
+
return f"An error occurred: {e}"
|
73 |
+
|
74 |
+
iface = gr.Interface(
|
75 |
+
fn=classify,
|
76 |
+
inputs="text",
|
77 |
+
outputs="text",
|
78 |
+
title="Waste Classifier",
|
79 |
+
description="Upload an image to classify and get disposal instructions."
|
80 |
+
)
|
81 |
+
iface.launch()
|