Spaces:
Runtime error
Runtime error
fix app
Browse files- app.py +35 -63
- config.yaml +0 -11
app.py
CHANGED
@@ -1,26 +1,29 @@
|
|
1 |
from typing import Any
|
2 |
import pytorch_lightning as pl
|
3 |
-
from torchvision.models import
|
4 |
import torch
|
5 |
from torch import nn
|
6 |
from torchvision import transforms
|
7 |
-
from torch.nn import functional as F
|
8 |
import yaml
|
9 |
from yaml.loader import SafeLoader
|
10 |
-
from PIL import Image
|
11 |
import gradio as gr
|
12 |
import os
|
13 |
|
|
|
14 |
class WeedModel(pl.LightningModule):
|
15 |
def __init__(self, params):
|
16 |
super().__init__()
|
17 |
self.params = params
|
18 |
-
|
19 |
model = self.params["model"]
|
20 |
|
21 |
-
if
|
22 |
-
if
|
23 |
-
|
|
|
|
|
|
|
|
|
24 |
num_ftrs = self.base_model.classifier[-1].in_features
|
25 |
self.base_model.classifier[-1] = nn.Linear(num_ftrs, self.params["n_class"])
|
26 |
|
@@ -31,41 +34,15 @@ class WeedModel(pl.LightningModule):
|
|
31 |
embedding = self.base_model(x)
|
32 |
return embedding
|
33 |
|
34 |
-
def
|
35 |
-
|
36 |
-
|
37 |
-
elif(self.params["optimizer"] == "SGD"):
|
38 |
-
optimizer = torch.optim.SGD(self.parameters(), lr=self.params["Lr"])
|
39 |
-
return optimizer
|
40 |
-
|
41 |
-
def training_step(self, train_batch, batch_idx):
|
42 |
-
x = train_batch["image"]
|
43 |
-
y = train_batch["label"]
|
44 |
-
|
45 |
-
y_hat = self(x)
|
46 |
-
loss = F.cross_entropy(y_hat, y)
|
47 |
-
self.log('metrics/batch/train_loss', loss, prog_bar=False)
|
48 |
-
|
49 |
-
preds = F.softmax(y_hat, dim=-1)
|
50 |
-
|
51 |
-
return loss
|
52 |
-
|
53 |
-
def validation_step(self, val_batch, batch_idx):
|
54 |
-
|
55 |
-
x = val_batch["image"]
|
56 |
-
y = val_batch["label"]
|
57 |
-
|
58 |
-
y_hat = self(x)
|
59 |
-
loss = F.cross_entropy(y_hat, y)
|
60 |
-
self.log('metrics/batch/val_loss', loss)
|
61 |
-
|
62 |
-
def predict_step(self, batch: Any, batch_idx: int=0, dataloader_idx: int = 0) -> Any:
|
63 |
y_hat = self(batch)
|
64 |
preds = torch.softmax(y_hat, dim=-1).tolist()
|
65 |
-
|
66 |
# preds = torch.argmax(preds, dim=-1)
|
67 |
return preds
|
68 |
-
|
69 |
|
70 |
def predict(image):
|
71 |
|
@@ -80,45 +57,40 @@ title = " AISeed AI Application Demo "
|
|
80 |
description = "# A Demo of Deep Learning for Weed Classification"
|
81 |
example_list = [["examples/" + example] for example in os.listdir("examples")]
|
82 |
|
83 |
-
with open("class_names.txt", "r", encoding=
|
84 |
class_names = f.read().splitlines()
|
85 |
-
|
86 |
with gr.Blocks() as demo:
|
87 |
demo.title = title
|
88 |
gr.Markdown(description)
|
89 |
with gr.Tabs():
|
90 |
-
with gr.TabItem("
|
91 |
with gr.Row():
|
92 |
with gr.Column():
|
93 |
-
im = gr.Image(type="pil", label="input image")
|
94 |
with gr.Column():
|
95 |
label_conv = gr.Label(label="Predictions", num_top_classes=4)
|
96 |
btn = gr.Button(value="predict")
|
97 |
btn.click(predict, inputs=im, outputs=[label_conv])
|
98 |
gr.Examples(examples=example_list, inputs=[im], outputs=[label_conv])
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
|
103 |
-
# capture = gr.Image(type="pil", label="output image")
|
104 |
-
with gr.Column():
|
105 |
-
label = gr.Label(label="Predictions", num_top_classes=4)
|
106 |
-
|
107 |
-
webcam.change(predict, inputs=webcam, outputs=[label])
|
108 |
-
|
109 |
-
|
110 |
-
if __name__ == '__main__':
|
111 |
-
with open('config.yaml') as f:
|
112 |
PARAMS = yaml.load(f, Loader=SafeLoader)
|
113 |
print(PARAMS)
|
114 |
-
model = WeedModel.load_from_checkpoint(
|
|
|
|
|
115 |
model.eval()
|
116 |
|
117 |
-
transform = transforms.Compose(
|
118 |
-
|
119 |
-
|
120 |
-
|
121 |
-
|
122 |
-
|
123 |
-
|
|
|
|
|
124 |
demo.launch()
|
|
|
1 |
from typing import Any
|
2 |
import pytorch_lightning as pl
|
3 |
+
from torchvision.models import efficientnet_v2_s, EfficientNet_V2_S_Weights
|
4 |
import torch
|
5 |
from torch import nn
|
6 |
from torchvision import transforms
|
|
|
7 |
import yaml
|
8 |
from yaml.loader import SafeLoader
|
|
|
9 |
import gradio as gr
|
10 |
import os
|
11 |
|
12 |
+
|
13 |
class WeedModel(pl.LightningModule):
|
14 |
def __init__(self, params):
|
15 |
super().__init__()
|
16 |
self.params = params
|
17 |
+
|
18 |
model = self.params["model"]
|
19 |
|
20 |
+
if model.lower() == "efficientnet":
|
21 |
+
if self.params["pretrained"]:
|
22 |
+
self.base_model = efficientnet_v2_s(
|
23 |
+
weights=EfficientNet_V2_S_Weights.IMAGENET1K_V1
|
24 |
+
)
|
25 |
+
else:
|
26 |
+
self.base_model = efficientnet_v2_s(weights=None)
|
27 |
num_ftrs = self.base_model.classifier[-1].in_features
|
28 |
self.base_model.classifier[-1] = nn.Linear(num_ftrs, self.params["n_class"])
|
29 |
|
|
|
34 |
embedding = self.base_model(x)
|
35 |
return embedding
|
36 |
|
37 |
+
def predict_step(
|
38 |
+
self, batch: Any, batch_idx: int = 0, dataloader_idx: int = 0
|
39 |
+
) -> Any:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
40 |
y_hat = self(batch)
|
41 |
preds = torch.softmax(y_hat, dim=-1).tolist()
|
42 |
+
|
43 |
# preds = torch.argmax(preds, dim=-1)
|
44 |
return preds
|
45 |
+
|
46 |
|
47 |
def predict(image):
|
48 |
|
|
|
57 |
description = "# A Demo of Deep Learning for Weed Classification"
|
58 |
example_list = [["examples/" + example] for example in os.listdir("examples")]
|
59 |
|
60 |
+
with open("class_names.txt", "r", encoding="utf-8") as f:
|
61 |
class_names = f.read().splitlines()
|
62 |
+
|
63 |
with gr.Blocks() as demo:
|
64 |
demo.title = title
|
65 |
gr.Markdown(description)
|
66 |
with gr.Tabs():
|
67 |
+
with gr.TabItem("Images"):
|
68 |
with gr.Row():
|
69 |
with gr.Column():
|
70 |
+
im = gr.Image(type="pil", label="input image", sources=["upload", "webcam"])
|
71 |
with gr.Column():
|
72 |
label_conv = gr.Label(label="Predictions", num_top_classes=4)
|
73 |
btn = gr.Button(value="predict")
|
74 |
btn.click(predict, inputs=im, outputs=[label_conv])
|
75 |
gr.Examples(examples=example_list, inputs=[im], outputs=[label_conv])
|
76 |
+
|
77 |
+
|
78 |
+
if __name__ == "__main__":
|
79 |
+
with open("config.yaml") as f:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
80 |
PARAMS = yaml.load(f, Loader=SafeLoader)
|
81 |
print(PARAMS)
|
82 |
+
model = WeedModel.load_from_checkpoint(
|
83 |
+
"model/epoch=08.ckpt", params=PARAMS, map_location=torch.device("cpu")
|
84 |
+
)
|
85 |
model.eval()
|
86 |
|
87 |
+
transform = transforms.Compose(
|
88 |
+
[
|
89 |
+
transforms.Resize(256),
|
90 |
+
transforms.CenterCrop(224),
|
91 |
+
transforms.ToTensor(),
|
92 |
+
# transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
|
93 |
+
]
|
94 |
+
)
|
95 |
+
|
96 |
demo.launch()
|
config.yaml
CHANGED
@@ -1,19 +1,8 @@
|
|
1 |
{
|
2 |
-
#Dataset
|
3 |
-
"train_path": "./train.txt",
|
4 |
-
"test_path": "./test.txt",
|
5 |
-
"val_path": "./val.txt",
|
6 |
-
"n_data": 1,
|
7 |
-
|
8 |
|
9 |
#Model
|
10 |
"model": "EfficientNet", # [Alexnet, VGG, GoogleNet, ResNet, DenseNet, MobileNet, SqueezeNet, ShuffleNet, EfficientNet, SE-ResNet (not available)]
|
11 |
"pretrained": True,
|
12 |
"n_class": 40,
|
13 |
|
14 |
-
#Training
|
15 |
-
"B_sz": 4,
|
16 |
-
"Lr": 0.001,
|
17 |
-
"Epoch": 5,
|
18 |
-
"optimizer": "Adam" #[Adam, SGD]
|
19 |
}
|
|
|
1 |
{
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
|
3 |
#Model
|
4 |
"model": "EfficientNet", # [Alexnet, VGG, GoogleNet, ResNet, DenseNet, MobileNet, SqueezeNet, ShuffleNet, EfficientNet, SE-ResNet (not available)]
|
5 |
"pretrained": True,
|
6 |
"n_class": 40,
|
7 |
|
|
|
|
|
|
|
|
|
|
|
8 |
}
|