Hugging Face's logo

Spaces:
abhiswain
/
Gradio-Hindi-Character-Recognition
Runtime error

App Files Community
abhiswain commited on
Commit
46643d9
•
1 Parent(s): 662fc83

Upload 16 files

Browse files
Files changed (17) hide show
  1. .gitattributes +1 -0
  2. assests/App.gif +3 -0
  3. models/best_digit_model.pt +3 -0
  4. models/best_vyanjan_model.pt +3 -0
  5. src/__pycache__/config.cpython-39.pyc +0 -0
  6. src/__pycache__/data.cpython-39.pyc +0 -0
  7. src/__pycache__/model.cpython-39.pyc +0 -0
  8. src/config.py +24 -0
  9. src/data.py +23 -0
  10. src/index_to_digit.json +12 -0
  11. src/index_to_vyanjan.json +38 -0
  12. src/model.py +53 -0
  13. src/results.txt +29 -0
  14. src/test.py +55 -0
  15. src/train.log +809 -0
  16. src/train.py +265 -0
  17. src/vyanjan_mapping.png +0 -0
.gitattributes CHANGED
@@ -32,3 +32,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
35
+ assests/App.gif filter=lfs diff=lfs merge=lfs -text
assests/App.gif ADDED

Git LFS Details

  • SHA256: eb6e2bdf9f438e28dc5b7bdb86da6fd0c49d36567c233c1094be0918da70a753
  • Pointer size: 132 Bytes
  • Size of remote file: 8.85 MB
models/best_digit_model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f1ef24cf9d1a08d869a836fe81a577e7e2c012a5949f8b9c79ea3022bd73f3e8
3
+ size 890647
models/best_vyanjan_model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:23298e51a89c2590b45ba0a64a0786d229ce9859dc7cd297d38d792bf2bb3226
3
+ size 3140503
src/__pycache__/config.cpython-39.pyc ADDED
Binary file (872 Bytes). View file
 
src/__pycache__/data.cpython-39.pyc ADDED
Binary file (694 Bytes). View file
 
src/__pycache__/model.cpython-39.pyc ADDED
Binary file (2.02 kB). View file
 
src/config.py ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import torch
2
+ from pathlib import Path
3
+
4
+ # Paths
5
+ BASE_PATH = Path(__file__).resolve().parents[1]
6
+ TRAIN_VYANJAN_PATH = BASE_PATH / "data" / "Train_vyanjan"
7
+ TEST_VYANJAN_PATH = BASE_PATH / "data" / "Test_vyanjan"
8
+ TRAIN_DIGIT_PATH = BASE_PATH / "data" / "Train_digits"
9
+ TEST_DIGIT_PATH = BASE_PATH / "data" / "Test_digits"
10
+ BEST_MODEL_VYANJAN = BASE_PATH / "models" / "best_vyanjan_model.pt"
11
+ BEST_MODEL_DIGIT = BASE_PATH / "models" / "best_digit_model.pt"
12
+ BEST_MODEL_PATH = ""
13
+ INDEX_DIGIT = BASE_PATH / "src" / "index_to_digit.json"
14
+ INDEX_VYNAJAN = BASE_PATH / "src" / "index_to_vyanjan.json"
15
+
16
+
17
+ # Hyperparameters
18
+ BATCH_SIZE = 32
19
+ EPOCHS = 10
20
+ LR = 1e-5
21
+
22
+ # Miscellanous
23
+ DEVICE = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")
24
+ INTERVAL = 100
src/data.py ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from torch.utils.data import DataLoader, random_split
2
+ from torchvision.datasets import ImageFolder
3
+ import torchvision.transforms as tfms
4
+ import torch
5
+
6
+ # the train & test transforms
7
+ transforms = {
8
+ "train": tfms.Compose(
9
+ [
10
+ tfms.PILToTensor(),
11
+ tfms.AutoAugment(tfms.AutoAugmentPolicy.IMAGENET),
12
+ tfms.ConvertImageDtype(torch.float),
13
+ tfms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
14
+ ]
15
+ ),
16
+ "test": tfms.Compose(
17
+ [
18
+ tfms.PILToTensor(),
19
+ tfms.ConvertImageDtype(torch.float),
20
+ tfms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
21
+ ]
22
+ ),
23
+ }
src/index_to_digit.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "0": "digit_0",
3
+ "1": "digit_1",
4
+ "2": "digit_2",
5
+ "3": "digit_3",
6
+ "4": "digit_4",
7
+ "5": "digit_5",
8
+ "6": "digit_6",
9
+ "7": "digit_7",
10
+ "8": "digit_8",
11
+ "9": "digit_9"
12
+ }
src/index_to_vyanjan.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "0": "character_10_nja",
3
+ "1": "character_11_Ta",
4
+ "2": "character_12_Tha",
5
+ "3": "character_13_Da",
6
+ "4": "character_14_Dha",
7
+ "5": "character_15_Na",
8
+ "6": "character_16_ta",
9
+ "7": "character_17_tha",
10
+ "8": "character_18_da",
11
+ "9": "character_19_dha",
12
+ "10": "character_1_ka",
13
+ "11": "character_20_na",
14
+ "12": "character_21_pa",
15
+ "13": "character_22_pha",
16
+ "14": "character_23_ba",
17
+ "15": "character_24_bha",
18
+ "16": "character_25_ma",
19
+ "17": "character_26_ya",
20
+ "18": "character_27_ra",
21
+ "19": "character_28_la",
22
+ "20": "character_29_wa",
23
+ "21": "character_2_Kha",
24
+ "22": "character_30_sha",
25
+ "23": "character_31_shha",
26
+ "24": "character_32_sa",
27
+ "25": "character_33_ha",
28
+ "26": "character_34_ksh",
29
+ "27": "character_35_tra",
30
+ "28": "character_36_gya",
31
+ "29": "character_3_Ga",
32
+ "30": "character_4_Gha",
33
+ "31": "character_5_nga",
34
+ "32": "character_6_cha",
35
+ "33": "character_7_chha",
36
+ "34": "character_8_ja",
37
+ "35": "character_9_jha"
38
+ }
src/model.py ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from torchvision.models import resnet18
2
+ import torch.nn as nn
3
+ import torch.nn.functional as F
4
+
5
+
6
+ def calculate_conv_output(IH, IW, KH, KW, P, S):
7
+ return ((IH - KH + 2 * P) / S + 1, (IW - KW + 2 * P) / S + 1)
8
+
9
+
10
+ class HNet(nn.Module):
11
+ def __init__(self, num_classes) -> None:
12
+ super().__init__()
13
+
14
+ # 32 x 32 x 3 => 28 x 28 x 16
15
+ self.conv1 = nn.Conv2d(3, 16, kernel_size=(5, 5))
16
+
17
+ # 28 x 28 x 16 => 26 x 26 x 32
18
+ self.conv2 = nn.Conv2d(16, 32, kernel_size=(3, 3))
19
+
20
+ # 26 x 26 x 32 => num_classes
21
+ self.fc1 = nn.Linear(26 * 26 * 32, num_classes)
22
+
23
+ self.dropout = nn.Dropout(p=0.5)
24
+
25
+ def forward(self, x):
26
+ x = F.relu(self.conv1(x))
27
+ x = F.relu(self.conv2(x))
28
+ x = x.view(-1, 26 * 26 * 32)
29
+ x = self.dropout(x)
30
+ x = self.fc1(x)
31
+
32
+ return x
33
+
34
+
35
+ class ResNet18(nn.Module):
36
+ def __init__(self, freeze=True, num_classes=10):
37
+ super(ResNet18, self).__init__()
38
+ self.resnet = resnet18(pretrained=True)
39
+
40
+ # freeze all layers if required
41
+ if freeze:
42
+ self.freeze_layers()
43
+
44
+ # new layers by default have requires_grad=True
45
+ self.resnet.fc = nn.Linear(512, num_classes)
46
+
47
+ def forward(self, x):
48
+ x = self.resnet(x)
49
+ return x
50
+
51
+ def freeze_layers(self):
52
+ for param in self.resnet.parameters():
53
+ param.requires_grad = False
src/results.txt ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ +-------+------------+-----------+----------+---------+
2
+ | Epoch | Train Loss | Train Acc | Val Loss | Val Acc |
3
+ +-------+------------+-----------+----------+---------+
4
+ | 1 | 2.215 | 22.868 | 2.096 | 51.029 |
5
+ | 2 | 1.36 | 61.706 | 0.953 | 72.706 |
6
+ | 3 | 0.847 | 73.765 | 0.666 | 80.529 |
7
+ | 4 | 0.663 | 79.559 | 0.536 | 83.471 |
8
+ | 5 | 0.535 | 83.684 | 0.418 | 88.441 |
9
+ | 6 | 0.394 | 88.037 | 0.334 | 90.588 |
10
+ | 7 | 0.302 | 91.037 | 0.275 | 91.706 |
11
+ | 8 | 0.256 | 92.493 | 0.239 | 93.824 |
12
+ | 9 | 0.218 | 93.449 | 0.204 | 94.412 |
13
+ | 10 | 0.205 | 93.787 | 0.179 | 95.088 |
14
+ | 11 | 0.178 | 94.588 | 0.182 | 95.176 |
15
+ | 12 | 0.172 | 95.103 | 0.172 | 95.118 |
16
+ | 13 | 0.155 | 95.36 | 0.152 | 95.853 |
17
+ | 14 | 0.146 | 95.61 | 0.151 | 95.853 |
18
+ | 15 | 0.15 | 95.699 | 0.153 | 95.441 |
19
+ | 16 | 0.132 | 96.022 | 0.145 | 96.0 |
20
+ | 17 | 0.127 | 96.022 | 0.147 | 96.235 |
21
+ | 18 | 0.126 | 96.191 | 0.137 | 96.441 |
22
+ | 19 | 0.126 | 96.309 | 0.168 | 95.618 |
23
+ | 20 | 0.116 | 96.434 | 0.135 | 96.647 |
24
+ | 21 | 0.116 | 96.603 | 0.145 | 96.353 |
25
+ | 22 | 0.107 | 96.75 | 0.127 | 96.853 |
26
+ | 23 | 0.107 | 96.904 | 0.127 | 96.559 |
27
+ | 24 | 0.104 | 96.853 | 0.137 | 96.353 |
28
+ | 25 | 0.101 | 97.037 | 0.121 | 96.559 |
29
+ +-------+------------+-----------+----------+---------+
src/test.py ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import torch
2
+ from torch.utils.data import DataLoader
3
+ from torchvision.datasets import ImageFolder
4
+ from data import transforms
5
+ from model import HNet, ResNet18
6
+ from tqdm import tqdm
7
+ import config as CFG
8
+ from argparse import ArgumentParser
9
+
10
+
11
+ def test(model_type):
12
+
13
+ model = None
14
+
15
+ if model_type == "digit":
16
+ test_ds = ImageFolder(root=CFG.TEST_DIGIT_PATH, transform=transforms["test"])
17
+ model = HNet(num_classes=10)
18
+ model.load_state_dict(torch.load(CFG.BEST_MODEL_DIGIT))
19
+ else:
20
+ test_ds = ImageFolder(root=CFG.TEST_VYANJAN_PATH, transform=transforms["test"])
21
+ model = HNet(num_classes=36)
22
+ model.load_state_dict(torch.load(CFG.BEST_MODEL_VYANJAN))
23
+
24
+ test_dl = DataLoader(test_ds, batch_size=CFG.BATCH_SIZE)
25
+
26
+ model.eval()
27
+
28
+ running_corrects = 0
29
+
30
+ with torch.no_grad():
31
+ for images, labels in tqdm(test_dl):
32
+
33
+ outputs = model(images)
34
+
35
+ _, preds = torch.max(outputs, 1)
36
+
37
+ running_corrects += torch.sum(preds == labels)
38
+
39
+ print(
40
+ f"Test Accuracy of [{model_type}] model: {round(running_corrects.item()/len(test_ds) * 100, 3)}%"
41
+ )
42
+
43
+
44
+ if __name__ == "__main__":
45
+ parser = ArgumentParser(description="Test model for Hindi Character Recognition")
46
+ parser.add_argument(
47
+ "--model_type",
48
+ type=str,
49
+ help="Type of model (vyanjan/digit)",
50
+ default="vyanjan",
51
+ )
52
+
53
+ args = parser.parse_args()
54
+
55
+ test(model_type=args.model_type)
src/train.log ADDED
@@ -0,0 +1,809 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2023-01-04 19:33:16,533 - INFO - Initialized Vyanjan model
2
+ 2023-01-04 19:33:16,941 - INFO -
3
+ Training details:
4
+ ------------------------
5
+ Model: HNet
6
+ Model Type: vyanjan
7
+ Epochs: 25
8
+ Optimizer: SGD
9
+ Loss: CrossEntropyLoss
10
+ Learning Rate: 1e-05
11
+ Learning Rate Scheduler: <torch.optim.lr_scheduler.CyclicLR object at 0x000001CF092DFBB0>
12
+ Batch Size: 32
13
+ Logging Interval: 100 batches
14
+ Train-dataset samples: 48960
15
+ Validation-dataset samples: 12240
16
+ -------------------------
17
+
18
+ 2023-01-04 19:33:16,941 - INFO - TRAIN phase
19
+ 2023-01-04 19:33:20,383 - INFO - Epoch 0 - TRAIN - Batch 0 - Loss = 3.615 | Accuracy = 0.0%
20
+ 2023-01-04 19:33:27,344 - INFO - Epoch 0 - TRAIN - Batch 100 - Loss = 3.59 | Accuracy = 3.125%
21
+ 2023-01-04 19:33:34,262 - INFO - Epoch 0 - TRAIN - Batch 200 - Loss = 3.56 | Accuracy = 0.0%
22
+ 2023-01-04 19:33:40,938 - INFO - Epoch 0 - TRAIN - Batch 300 - Loss = 3.565 | Accuracy = 3.125%
23
+ 2023-01-04 19:33:48,369 - INFO - Epoch 0 - TRAIN - Batch 400 - Loss = 3.544 | Accuracy = 6.25%
24
+ 2023-01-04 19:33:54,067 - INFO - Epoch 0 - TRAIN - Batch 500 - Loss = 3.563 | Accuracy = 6.25%
25
+ 2023-01-04 19:33:59,775 - INFO - Epoch 0 - TRAIN - Batch 600 - Loss = 3.501 | Accuracy = 6.25%
26
+ 2023-01-04 19:34:05,646 - INFO - Epoch 0 - TRAIN - Batch 700 - Loss = 3.578 | Accuracy = 6.25%
27
+ 2023-01-04 19:34:11,282 - INFO - Epoch 0 - TRAIN - Batch 800 - Loss = 3.557 | Accuracy = 6.25%
28
+ 2023-01-04 19:34:16,849 - INFO - Epoch 0 - TRAIN - Batch 900 - Loss = 3.431 | Accuracy = 15.625%
29
+ 2023-01-04 19:34:22,412 - INFO - Epoch 0 - TRAIN - Batch 1000 - Loss = 3.549 | Accuracy = 0.0%
30
+ 2023-01-04 19:34:28,909 - INFO - Epoch 0 - TRAIN - Batch 1100 - Loss = 3.451 | Accuracy = 12.5%
31
+ 2023-01-04 19:34:34,880 - INFO - Epoch 0 - TRAIN - Batch 1200 - Loss = 3.462 | Accuracy = 15.625%
32
+ 2023-01-04 19:34:40,872 - INFO - Epoch 0 - TRAIN - Batch 1300 - Loss = 3.484 | Accuracy = 3.125%
33
+ 2023-01-04 19:34:46,838 - INFO - Epoch 0 - TRAIN - Batch 1400 - Loss = 3.358 | Accuracy = 21.875%
34
+ 2023-01-04 19:34:52,829 - INFO - Epoch 0 - TRAIN - Batch 1500 - Loss = 3.523 | Accuracy = 6.25%
35
+ 2023-01-04 19:34:54,503 - INFO - VAL phase
36
+ 2023-01-04 19:34:54,555 - INFO - Epoch 0 - VAL - Batch 0 - Loss = 3.409 | Accuracy = 18.75%
37
+ 2023-01-04 19:34:59,772 - INFO - Epoch 0 - VAL - Batch 100 - Loss = 3.394 | Accuracy = 25.0%
38
+ 2023-01-04 19:35:05,112 - INFO - Epoch 0 - VAL - Batch 200 - Loss = 3.371 | Accuracy = 25.0%
39
+ 2023-01-04 19:35:10,557 - INFO - Epoch 0 - VAL - Batch 300 - Loss = 3.414 | Accuracy = 21.875%
40
+ 2023-01-04 19:35:15,491 - INFO - TRAIN phase
41
+ 2023-01-04 19:35:15,579 - INFO - Epoch 1 - TRAIN - Batch 0 - Loss = 3.438 | Accuracy = 6.25%
42
+ 2023-01-04 19:35:21,124 - INFO - Epoch 1 - TRAIN - Batch 100 - Loss = 3.325 | Accuracy = 15.625%
43
+ 2023-01-04 19:35:26,359 - INFO - Epoch 1 - TRAIN - Batch 200 - Loss = 3.205 | Accuracy = 18.75%
44
+ 2023-01-04 19:35:31,512 - INFO - Epoch 1 - TRAIN - Batch 300 - Loss = 2.752 | Accuracy = 31.25%
45
+ 2023-01-04 19:35:36,782 - INFO - Epoch 1 - TRAIN - Batch 400 - Loss = 3.006 | Accuracy = 15.625%
46
+ 2023-01-04 19:35:42,040 - INFO - Epoch 1 - TRAIN - Batch 500 - Loss = 3.029 | Accuracy = 31.25%
47
+ 2023-01-04 19:35:47,394 - INFO - Epoch 1 - TRAIN - Batch 600 - Loss = 2.374 | Accuracy = 31.25%
48
+ 2023-01-04 19:35:52,839 - INFO - Epoch 1 - TRAIN - Batch 700 - Loss = 2.599 | Accuracy = 40.625%
49
+ 2023-01-04 19:35:58,317 - INFO - Epoch 1 - TRAIN - Batch 800 - Loss = 2.36 | Accuracy = 43.75%
50
+ 2023-01-04 19:36:04,365 - INFO - Epoch 1 - TRAIN - Batch 900 - Loss = 2.073 | Accuracy = 53.125%
51
+ 2023-01-04 19:36:10,584 - INFO - Epoch 1 - TRAIN - Batch 1000 - Loss = 2.473 | Accuracy = 40.625%
52
+ 2023-01-04 19:36:16,198 - INFO - Epoch 1 - TRAIN - Batch 1100 - Loss = 2.115 | Accuracy = 37.5%
53
+ 2023-01-04 19:36:21,774 - INFO - Epoch 1 - TRAIN - Batch 1200 - Loss = 2.328 | Accuracy = 25.0%
54
+ 2023-01-04 19:36:27,339 - INFO - Epoch 1 - TRAIN - Batch 1300 - Loss = 2.014 | Accuracy = 50.0%
55
+ 2023-01-04 19:36:33,158 - INFO - Epoch 1 - TRAIN - Batch 1400 - Loss = 2.424 | Accuracy = 37.5%
56
+ 2023-01-04 19:36:38,756 - INFO - Epoch 1 - TRAIN - Batch 1500 - Loss = 2.439 | Accuracy = 34.375%
57
+ 2023-01-04 19:36:40,398 - INFO - VAL phase
58
+ 2023-01-04 19:36:40,460 - INFO - Epoch 1 - VAL - Batch 0 - Loss = 1.982 | Accuracy = 56.25%
59
+ 2023-01-04 19:36:45,360 - INFO - Epoch 1 - VAL - Batch 100 - Loss = 2.075 | Accuracy = 53.125%
60
+ 2023-01-04 19:36:50,468 - INFO - Epoch 1 - VAL - Batch 200 - Loss = 2.045 | Accuracy = 43.75%
61
+ 2023-01-04 19:36:55,596 - INFO - Epoch 1 - VAL - Batch 300 - Loss = 1.821 | Accuracy = 56.25%
62
+ 2023-01-04 19:37:00,523 - INFO - TRAIN phase
63
+ 2023-01-04 19:37:00,617 - INFO - Epoch 2 - TRAIN - Batch 0 - Loss = 2.278 | Accuracy = 40.625%
64
+ 2023-01-04 19:37:05,838 - INFO - Epoch 2 - TRAIN - Batch 100 - Loss = 2.042 | Accuracy = 40.625%
65
+ 2023-01-04 19:37:11,034 - INFO - Epoch 2 - TRAIN - Batch 200 - Loss = 2.163 | Accuracy = 43.75%
66
+ 2023-01-04 19:37:16,438 - INFO - Epoch 2 - TRAIN - Batch 300 - Loss = 2.02 | Accuracy = 46.875%
67
+ 2023-01-04 19:37:21,768 - INFO - Epoch 2 - TRAIN - Batch 400 - Loss = 2.326 | Accuracy = 37.5%
68
+ 2023-01-04 19:37:27,130 - INFO - Epoch 2 - TRAIN - Batch 500 - Loss = 1.813 | Accuracy = 46.875%
69
+ 2023-01-04 19:37:32,821 - INFO - Epoch 2 - TRAIN - Batch 600 - Loss = 2.403 | Accuracy = 31.25%
70
+ 2023-01-04 19:37:38,150 - INFO - Epoch 2 - TRAIN - Batch 700 - Loss = 2.047 | Accuracy = 34.375%
71
+ 2023-01-04 19:37:43,654 - INFO - Epoch 2 - TRAIN - Batch 800 - Loss = 2.25 | Accuracy = 34.375%
72
+ 2023-01-04 19:37:50,078 - INFO - Epoch 2 - TRAIN - Batch 900 - Loss = 2.337 | Accuracy = 53.125%
73
+ 2023-01-04 19:37:56,346 - INFO - Epoch 2 - TRAIN - Batch 1000 - Loss = 1.726 | Accuracy = 56.25%
74
+ 2023-01-04 19:38:02,813 - INFO - Epoch 2 - TRAIN - Batch 1100 - Loss = 1.781 | Accuracy = 46.875%
75
+ 2023-01-04 19:38:09,074 - INFO - Epoch 2 - TRAIN - Batch 1200 - Loss = 1.839 | Accuracy = 50.0%
76
+ 2023-01-04 19:38:15,410 - INFO - Epoch 2 - TRAIN - Batch 1300 - Loss = 1.916 | Accuracy = 37.5%
77
+ 2023-01-04 19:38:21,859 - INFO - Epoch 2 - TRAIN - Batch 1400 - Loss = 1.622 | Accuracy = 56.25%
78
+ 2023-01-04 19:38:28,203 - INFO - Epoch 2 - TRAIN - Batch 1500 - Loss = 1.749 | Accuracy = 56.25%
79
+ 2023-01-04 19:38:30,047 - INFO - VAL phase
80
+ 2023-01-04 19:38:30,095 - INFO - Epoch 2 - VAL - Batch 0 - Loss = 1.096 | Accuracy = 68.75%
81
+ 2023-01-04 19:38:35,774 - INFO - Epoch 2 - VAL - Batch 100 - Loss = 1.424 | Accuracy = 65.625%
82
+ 2023-01-04 19:38:41,225 - INFO - Epoch 2 - VAL - Batch 200 - Loss = 1.408 | Accuracy = 65.625%
83
+ 2023-01-04 19:38:46,612 - INFO - Epoch 2 - VAL - Batch 300 - Loss = 1.352 | Accuracy = 65.625%
84
+ 2023-01-04 19:38:50,972 - INFO - TRAIN phase
85
+ 2023-01-04 19:38:51,055 - INFO - Epoch 3 - TRAIN - Batch 0 - Loss = 1.808 | Accuracy = 62.5%
86
+ 2023-01-04 19:38:57,388 - INFO - Epoch 3 - TRAIN - Batch 100 - Loss = 1.563 | Accuracy = 53.125%
87
+ 2023-01-04 19:39:03,760 - INFO - Epoch 3 - TRAIN - Batch 200 - Loss = 1.307 | Accuracy = 68.75%
88
+ 2023-01-04 19:39:10,061 - INFO - Epoch 3 - TRAIN - Batch 300 - Loss = 1.568 | Accuracy = 71.875%
89
+ 2023-01-04 19:39:16,445 - INFO - Epoch 3 - TRAIN - Batch 400 - Loss = 1.469 | Accuracy = 53.125%
90
+ 2023-01-04 19:39:22,819 - INFO - Epoch 3 - TRAIN - Batch 500 - Loss = 1.649 | Accuracy = 53.125%
91
+ 2023-01-04 19:39:29,127 - INFO - Epoch 3 - TRAIN - Batch 600 - Loss = 1.491 | Accuracy = 53.125%
92
+ 2023-01-04 19:39:35,415 - INFO - Epoch 3 - TRAIN - Batch 700 - Loss = 1.361 | Accuracy = 59.375%
93
+ 2023-01-04 19:39:41,153 - INFO - Epoch 3 - TRAIN - Batch 800 - Loss = 1.33 | Accuracy = 56.25%
94
+ 2023-01-04 19:39:46,749 - INFO - Epoch 3 - TRAIN - Batch 900 - Loss = 1.02 | Accuracy = 81.25%
95
+ 2023-01-04 19:39:52,384 - INFO - Epoch 3 - TRAIN - Batch 1000 - Loss = 1.58 | Accuracy = 59.375%
96
+ 2023-01-04 19:39:57,956 - INFO - Epoch 3 - TRAIN - Batch 1100 - Loss = 2.078 | Accuracy = 59.375%
97
+ 2023-01-04 19:40:03,543 - INFO - Epoch 3 - TRAIN - Batch 1200 - Loss = 1.424 | Accuracy = 53.125%
98
+ 2023-01-04 19:40:09,226 - INFO - Epoch 3 - TRAIN - Batch 1300 - Loss = 1.392 | Accuracy = 59.375%
99
+ 2023-01-04 19:40:14,731 - INFO - Epoch 3 - TRAIN - Batch 1400 - Loss = 0.968 | Accuracy = 75.0%
100
+ 2023-01-04 19:40:19,920 - INFO - Epoch 3 - TRAIN - Batch 1500 - Loss = 1.25 | Accuracy = 56.25%
101
+ 2023-01-04 19:40:21,480 - INFO - VAL phase
102
+ 2023-01-04 19:40:21,536 - INFO - Epoch 3 - VAL - Batch 0 - Loss = 0.921 | Accuracy = 75.0%
103
+ 2023-01-04 19:40:26,390 - INFO - Epoch 3 - VAL - Batch 100 - Loss = 0.985 | Accuracy = 71.875%
104
+ 2023-01-04 19:40:31,189 - INFO - Epoch 3 - VAL - Batch 200 - Loss = 0.977 | Accuracy = 68.75%
105
+ 2023-01-04 19:40:36,009 - INFO - Epoch 3 - VAL - Batch 300 - Loss = 1.465 | Accuracy = 71.875%
106
+ 2023-01-04 19:40:40,481 - INFO - TRAIN phase
107
+ 2023-01-04 19:40:40,546 - INFO - Epoch 4 - TRAIN - Batch 0 - Loss = 1.262 | Accuracy = 62.5%
108
+ 2023-01-04 19:40:45,767 - INFO - Epoch 4 - TRAIN - Batch 100 - Loss = 1.136 | Accuracy = 68.75%
109
+ 2023-01-04 19:40:50,935 - INFO - Epoch 4 - TRAIN - Batch 200 - Loss = 1.134 | Accuracy = 75.0%
110
+ 2023-01-04 19:40:56,294 - INFO - Epoch 4 - TRAIN - Batch 300 - Loss = 0.741 | Accuracy = 78.125%
111
+ 2023-01-04 19:41:02,639 - INFO - Epoch 4 - TRAIN - Batch 400 - Loss = 0.86 | Accuracy = 68.75%
112
+ 2023-01-04 19:41:09,016 - INFO - Epoch 4 - TRAIN - Batch 500 - Loss = 1.078 | Accuracy = 75.0%
113
+ 2023-01-04 19:41:15,273 - INFO - Epoch 4 - TRAIN - Batch 600 - Loss = 1.392 | Accuracy = 65.625%
114
+ 2023-01-04 19:41:21,554 - INFO - Epoch 4 - TRAIN - Batch 700 - Loss = 1.31 | Accuracy = 62.5%
115
+ 2023-01-04 19:41:27,844 - INFO - Epoch 4 - TRAIN - Batch 800 - Loss = 1.666 | Accuracy = 53.125%
116
+ 2023-01-04 19:41:34,187 - INFO - Epoch 4 - TRAIN - Batch 900 - Loss = 1.209 | Accuracy = 65.625%
117
+ 2023-01-04 19:41:39,966 - INFO - Epoch 4 - TRAIN - Batch 1000 - Loss = 0.648 | Accuracy = 78.125%
118
+ 2023-01-04 19:41:45,129 - INFO - Epoch 4 - TRAIN - Batch 1100 - Loss = 1.591 | Accuracy = 65.625%
119
+ 2023-01-04 19:41:50,242 - INFO - Epoch 4 - TRAIN - Batch 1200 - Loss = 1.629 | Accuracy = 62.5%
120
+ 2023-01-04 19:41:55,994 - INFO - Epoch 4 - TRAIN - Batch 1300 - Loss = 0.962 | Accuracy = 71.875%
121
+ 2023-01-04 19:42:01,280 - INFO - Epoch 4 - TRAIN - Batch 1400 - Loss = 0.659 | Accuracy = 81.25%
122
+ 2023-01-04 19:42:06,399 - INFO - Epoch 4 - TRAIN - Batch 1500 - Loss = 1.094 | Accuracy = 75.0%
123
+ 2023-01-04 19:42:07,856 - INFO - VAL phase
124
+ 2023-01-04 19:42:07,910 - INFO - Epoch 4 - VAL - Batch 0 - Loss = 0.844 | Accuracy = 78.125%
125
+ 2023-01-04 19:42:12,442 - INFO - Epoch 4 - VAL - Batch 100 - Loss = 0.505 | Accuracy = 90.625%
126
+ 2023-01-04 19:42:17,030 - INFO - Epoch 4 - VAL - Batch 200 - Loss = 0.777 | Accuracy = 84.375%
127
+ 2023-01-04 19:42:21,709 - INFO - Epoch 4 - VAL - Batch 300 - Loss = 1.008 | Accuracy = 75.0%
128
+ 2023-01-04 19:42:25,396 - INFO - TRAIN phase
129
+ 2023-01-04 19:42:25,452 - INFO - Epoch 5 - TRAIN - Batch 0 - Loss = 1.183 | Accuracy = 75.0%
130
+ 2023-01-04 19:42:31,740 - INFO - Epoch 5 - TRAIN - Batch 100 - Loss = 1.097 | Accuracy = 59.375%
131
+ 2023-01-04 19:42:38,456 - INFO - Epoch 5 - TRAIN - Batch 200 - Loss = 0.829 | Accuracy = 75.0%
132
+ 2023-01-04 19:42:43,478 - INFO - Epoch 5 - TRAIN - Batch 300 - Loss = 0.677 | Accuracy = 81.25%
133
+ 2023-01-04 19:42:49,302 - INFO - Epoch 5 - TRAIN - Batch 400 - Loss = 0.627 | Accuracy = 75.0%
134
+ 2023-01-04 19:42:55,197 - INFO - Epoch 5 - TRAIN - Batch 500 - Loss = 0.948 | Accuracy = 68.75%
135
+ 2023-01-04 19:43:01,115 - INFO - Epoch 5 - TRAIN - Batch 600 - Loss = 0.95 | Accuracy = 78.125%
136
+ 2023-01-04 19:43:06,903 - INFO - Epoch 5 - TRAIN - Batch 700 - Loss = 0.954 | Accuracy = 68.75%
137
+ 2023-01-04 19:43:12,785 - INFO - Epoch 5 - TRAIN - Batch 800 - Loss = 1.184 | Accuracy = 65.625%
138
+ 2023-01-04 19:43:18,643 - INFO - Epoch 5 - TRAIN - Batch 900 - Loss = 0.988 | Accuracy = 71.875%
139
+ 2023-01-04 19:43:24,429 - INFO - Epoch 5 - TRAIN - Batch 1000 - Loss = 1.183 | Accuracy = 71.875%
140
+ 2023-01-04 19:43:30,191 - INFO - Epoch 5 - TRAIN - Batch 1100 - Loss = 0.93 | Accuracy = 78.125%
141
+ 2023-01-04 19:43:35,958 - INFO - Epoch 5 - TRAIN - Batch 1200 - Loss = 0.962 | Accuracy = 78.125%
142
+ 2023-01-04 19:43:41,192 - INFO - Epoch 5 - TRAIN - Batch 1300 - Loss = 0.708 | Accuracy = 84.375%
143
+ 2023-01-04 19:43:46,332 - INFO - Epoch 5 - TRAIN - Batch 1400 - Loss = 1.221 | Accuracy = 62.5%
144
+ 2023-01-04 19:43:51,431 - INFO - Epoch 5 - TRAIN - Batch 1500 - Loss = 0.544 | Accuracy = 90.625%
145
+ 2023-01-04 19:43:52,925 - INFO - VAL phase
146
+ 2023-01-04 19:43:52,980 - INFO - Epoch 5 - VAL - Batch 0 - Loss = 0.754 | Accuracy = 78.125%
147
+ 2023-01-04 19:43:57,552 - INFO - Epoch 5 - VAL - Batch 100 - Loss = 0.847 | Accuracy = 81.25%
148
+ 2023-01-04 19:44:02,196 - INFO - Epoch 5 - VAL - Batch 200 - Loss = 0.494 | Accuracy = 90.625%
149
+ 2023-01-04 19:44:06,815 - INFO - Epoch 5 - VAL - Batch 300 - Loss = 0.834 | Accuracy = 81.25%
150
+ 2023-01-04 19:44:10,651 - INFO - TRAIN phase
151
+ 2023-01-04 19:44:10,715 - INFO - Epoch 6 - TRAIN - Batch 0 - Loss = 1.152 | Accuracy = 75.0%
152
+ 2023-01-04 19:44:15,833 - INFO - Epoch 6 - TRAIN - Batch 100 - Loss = 0.881 | Accuracy = 78.125%
153
+ 2023-01-04 19:44:21,132 - INFO - Epoch 6 - TRAIN - Batch 200 - Loss = 0.746 | Accuracy = 87.5%
154
+ 2023-01-04 19:44:26,648 - INFO - Epoch 6 - TRAIN - Batch 300 - Loss = 1.079 | Accuracy = 68.75%
155
+ 2023-01-04 19:44:31,929 - INFO - Epoch 6 - TRAIN - Batch 400 - Loss = 0.847 | Accuracy = 71.875%
156
+ 2023-01-04 19:44:36,890 - INFO - Epoch 6 - TRAIN - Batch 500 - Loss = 1.356 | Accuracy = 71.875%
157
+ 2023-01-04 19:44:42,116 - INFO - Epoch 6 - TRAIN - Batch 600 - Loss = 0.871 | Accuracy = 81.25%
158
+ 2023-01-04 19:44:47,409 - INFO - Epoch 6 - TRAIN - Batch 700 - Loss = 0.86 | Accuracy = 75.0%
159
+ 2023-01-04 19:44:52,559 - INFO - Epoch 6 - TRAIN - Batch 800 - Loss = 1.288 | Accuracy = 75.0%
160
+ 2023-01-04 19:44:57,801 - INFO - Epoch 6 - TRAIN - Batch 900 - Loss = 0.81 | Accuracy = 75.0%
161
+ 2023-01-04 19:45:02,945 - INFO - Epoch 6 - TRAIN - Batch 1000 - Loss = 0.384 | Accuracy = 93.75%
162
+ 2023-01-04 19:45:08,129 - INFO - Epoch 6 - TRAIN - Batch 1100 - Loss = 0.543 | Accuracy = 84.375%
163
+ 2023-01-04 19:45:13,513 - INFO - Epoch 6 - TRAIN - Batch 1200 - Loss = 0.771 | Accuracy = 75.0%
164
+ 2023-01-04 19:45:18,258 - INFO - Epoch 6 - TRAIN - Batch 1300 - Loss = 0.513 | Accuracy = 81.25%
165
+ 2023-01-04 19:45:23,168 - INFO - Epoch 6 - TRAIN - Batch 1400 - Loss = 0.529 | Accuracy = 81.25%
166
+ 2023-01-04 19:45:27,974 - INFO - Epoch 6 - TRAIN - Batch 1500 - Loss = 0.933 | Accuracy = 62.5%
167
+ 2023-01-04 19:45:29,389 - INFO - VAL phase
168
+ 2023-01-04 19:45:29,437 - INFO - Epoch 6 - VAL - Batch 0 - Loss = 0.462 | Accuracy = 87.5%
169
+ 2023-01-04 19:45:33,999 - INFO - Epoch 6 - VAL - Batch 100 - Loss = 0.638 | Accuracy = 84.375%
170
+ 2023-01-04 19:45:38,585 - INFO - Epoch 6 - VAL - Batch 200 - Loss = 0.694 | Accuracy = 87.5%
171
+ 2023-01-04 19:45:43,069 - INFO - Epoch 6 - VAL - Batch 300 - Loss = 1.131 | Accuracy = 78.125%
172
+ 2023-01-04 19:45:46,779 - INFO - TRAIN phase
173
+ 2023-01-04 19:45:46,844 - INFO - Epoch 7 - TRAIN - Batch 0 - Loss = 0.778 | Accuracy = 81.25%
174
+ 2023-01-04 19:45:51,791 - INFO - Epoch 7 - TRAIN - Batch 100 - Loss = 0.566 | Accuracy = 81.25%
175
+ 2023-01-04 19:45:56,791 - INFO - Epoch 7 - TRAIN - Batch 200 - Loss = 0.641 | Accuracy = 84.375%
176
+ 2023-01-04 19:46:02,205 - INFO - Epoch 7 - TRAIN - Batch 300 - Loss = 0.878 | Accuracy = 71.875%
177
+ 2023-01-04 19:46:07,606 - INFO - Epoch 7 - TRAIN - Batch 400 - Loss = 0.564 | Accuracy = 87.5%
178
+ 2023-01-04 19:46:12,888 - INFO - Epoch 7 - TRAIN - Batch 500 - Loss = 0.436 | Accuracy = 90.625%
179
+ 2023-01-04 19:46:18,108 - INFO - Epoch 7 - TRAIN - Batch 600 - Loss = 0.62 | Accuracy = 75.0%
180
+ 2023-01-04 19:46:23,226 - INFO - Epoch 7 - TRAIN - Batch 700 - Loss = 0.852 | Accuracy = 71.875%
181
+ 2023-01-04 19:46:28,410 - INFO - Epoch 7 - TRAIN - Batch 800 - Loss = 0.644 | Accuracy = 84.375%
182
+ 2023-01-04 19:46:33,702 - INFO - Epoch 7 - TRAIN - Batch 900 - Loss = 0.578 | Accuracy = 90.625%
183
+ 2023-01-04 19:46:38,817 - INFO - Epoch 7 - TRAIN - Batch 1000 - Loss = 0.489 | Accuracy = 87.5%
184
+ 2023-01-04 19:46:43,961 - INFO - Epoch 7 - TRAIN - Batch 1100 - Loss = 0.793 | Accuracy = 81.25%
185
+ 2023-01-04 19:46:49,276 - INFO - Epoch 7 - TRAIN - Batch 1200 - Loss = 0.691 | Accuracy = 75.0%
186
+ 2023-01-04 19:46:55,296 - INFO - Epoch 7 - TRAIN - Batch 1300 - Loss = 0.893 | Accuracy = 75.0%
187
+ 2023-01-04 19:47:01,138 - INFO - Epoch 7 - TRAIN - Batch 1400 - Loss = 0.585 | Accuracy = 87.5%
188
+ 2023-01-04 19:47:07,054 - INFO - Epoch 7 - TRAIN - Batch 1500 - Loss = 0.685 | Accuracy = 78.125%
189
+ 2023-01-04 19:47:08,768 - INFO - VAL phase
190
+ 2023-01-04 19:47:08,817 - INFO - Epoch 7 - VAL - Batch 0 - Loss = 0.43 | Accuracy = 90.625%
191
+ 2023-01-04 19:47:13,632 - INFO - Epoch 7 - VAL - Batch 100 - Loss = 0.309 | Accuracy = 84.375%
192
+ 2023-01-04 19:47:18,430 - INFO - Epoch 7 - VAL - Batch 200 - Loss = 0.922 | Accuracy = 78.125%
193
+ 2023-01-04 19:47:23,395 - INFO - Epoch 7 - VAL - Batch 300 - Loss = 1.166 | Accuracy = 78.125%
194
+ 2023-01-04 19:47:27,289 - INFO - TRAIN phase
195
+ 2023-01-04 19:47:27,356 - INFO - Epoch 8 - TRAIN - Batch 0 - Loss = 0.755 | Accuracy = 75.0%
196
+ 2023-01-04 19:47:33,255 - INFO - Epoch 8 - TRAIN - Batch 100 - Loss = 0.425 | Accuracy = 87.5%
197
+ 2023-01-04 19:47:38,904 - INFO - Epoch 8 - TRAIN - Batch 200 - Loss = 1.032 | Accuracy = 78.125%
198
+ 2023-01-04 19:47:43,846 - INFO - Epoch 8 - TRAIN - Batch 300 - Loss = 0.631 | Accuracy = 90.625%
199
+ 2023-01-04 19:47:48,774 - INFO - Epoch 8 - TRAIN - Batch 400 - Loss = 0.156 | Accuracy = 96.875%
200
+ 2023-01-04 19:47:53,548 - INFO - Epoch 8 - TRAIN - Batch 500 - Loss = 0.682 | Accuracy = 84.375%
201
+ 2023-01-04 19:47:58,456 - INFO - Epoch 8 - TRAIN - Batch 600 - Loss = 0.715 | Accuracy = 81.25%
202
+ 2023-01-04 19:48:03,401 - INFO - Epoch 8 - TRAIN - Batch 700 - Loss = 0.565 | Accuracy = 84.375%
203
+ 2023-01-04 19:48:08,243 - INFO - Epoch 8 - TRAIN - Batch 800 - Loss = 0.678 | Accuracy = 81.25%
204
+ 2023-01-04 19:48:13,054 - INFO - Epoch 8 - TRAIN - Batch 900 - Loss = 0.655 | Accuracy = 84.375%
205
+ 2023-01-04 19:48:17,810 - INFO - Epoch 8 - TRAIN - Batch 1000 - Loss = 0.886 | Accuracy = 78.125%
206
+ 2023-01-04 19:48:22,705 - INFO - Epoch 8 - TRAIN - Batch 1100 - Loss = 0.575 | Accuracy = 81.25%
207
+ 2023-01-04 19:48:30,540 - INFO - Epoch 8 - TRAIN - Batch 1200 - Loss = 0.305 | Accuracy = 93.75%
208
+ 2023-01-04 19:48:36,577 - INFO - Epoch 8 - TRAIN - Batch 1300 - Loss = 0.483 | Accuracy = 87.5%
209
+ 2023-01-04 19:48:42,431 - INFO - Epoch 8 - TRAIN - Batch 1400 - Loss = 0.742 | Accuracy = 75.0%
210
+ 2023-01-04 19:48:48,327 - INFO - Epoch 8 - TRAIN - Batch 1500 - Loss = 0.596 | Accuracy = 87.5%
211
+ 2023-01-04 19:48:50,041 - INFO - VAL phase
212
+ 2023-01-04 19:48:50,099 - INFO - Epoch 8 - VAL - Batch 0 - Loss = 0.393 | Accuracy = 87.5%
213
+ 2023-01-04 19:48:54,842 - INFO - Epoch 8 - VAL - Batch 100 - Loss = 0.592 | Accuracy = 87.5%
214
+ 2023-01-04 19:48:59,673 - INFO - Epoch 8 - VAL - Batch 200 - Loss = 0.562 | Accuracy = 75.0%
215
+ 2023-01-04 19:49:04,524 - INFO - Epoch 8 - VAL - Batch 300 - Loss = 0.664 | Accuracy = 81.25%
216
+ 2023-01-04 19:49:08,402 - INFO - TRAIN phase
217
+ 2023-01-04 19:49:08,474 - INFO - Epoch 9 - TRAIN - Batch 0 - Loss = 0.953 | Accuracy = 75.0%
218
+ 2023-01-04 19:49:14,483 - INFO - Epoch 9 - TRAIN - Batch 100 - Loss = 0.654 | Accuracy = 78.125%
219
+ 2023-01-04 19:49:19,713 - INFO - Epoch 9 - TRAIN - Batch 200 - Loss = 0.553 | Accuracy = 81.25%
220
+ 2023-01-04 19:49:24,910 - INFO - Epoch 9 - TRAIN - Batch 300 - Loss = 0.626 | Accuracy = 75.0%
221
+ 2023-01-04 19:49:30,051 - INFO - Epoch 9 - TRAIN - Batch 400 - Loss = 0.567 | Accuracy = 81.25%
222
+ 2023-01-04 19:49:35,285 - INFO - Epoch 9 - TRAIN - Batch 500 - Loss = 0.393 | Accuracy = 84.375%
223
+ 2023-01-04 19:49:40,450 - INFO - Epoch 9 - TRAIN - Batch 600 - Loss = 0.474 | Accuracy = 87.5%
224
+ 2023-01-04 19:49:45,636 - INFO - Epoch 9 - TRAIN - Batch 700 - Loss = 0.785 | Accuracy = 81.25%
225
+ 2023-01-04 19:49:50,861 - INFO - Epoch 9 - TRAIN - Batch 800 - Loss = 0.33 | Accuracy = 90.625%
226
+ 2023-01-04 19:49:56,072 - INFO - Epoch 9 - TRAIN - Batch 900 - Loss = 0.316 | Accuracy = 93.75%
227
+ 2023-01-04 19:50:01,372 - INFO - Epoch 9 - TRAIN - Batch 1000 - Loss = 0.602 | Accuracy = 78.125%
228
+ 2023-01-04 19:50:07,322 - INFO - Epoch 9 - TRAIN - Batch 1100 - Loss = 0.899 | Accuracy = 78.125%
229
+ 2023-01-04 19:50:13,058 - INFO - Epoch 9 - TRAIN - Batch 1200 - Loss = 0.701 | Accuracy = 75.0%
230
+ 2023-01-04 19:50:18,940 - INFO - Epoch 9 - TRAIN - Batch 1300 - Loss = 0.694 | Accuracy = 75.0%
231
+ 2023-01-04 19:50:24,830 - INFO - Epoch 9 - TRAIN - Batch 1400 - Loss = 0.567 | Accuracy = 81.25%
232
+ 2023-01-04 19:50:30,526 - INFO - Epoch 9 - TRAIN - Batch 1500 - Loss = 0.472 | Accuracy = 87.5%
233
+ 2023-01-04 19:50:32,252 - INFO - VAL phase
234
+ 2023-01-04 19:50:32,303 - INFO - Epoch 9 - VAL - Batch 0 - Loss = 0.363 | Accuracy = 90.625%
235
+ 2023-01-04 19:50:37,046 - INFO - Epoch 9 - VAL - Batch 100 - Loss = 0.347 | Accuracy = 90.625%
236
+ 2023-01-04 19:50:41,807 - INFO - Epoch 9 - VAL - Batch 200 - Loss = 0.56 | Accuracy = 87.5%
237
+ 2023-01-04 19:50:46,613 - INFO - Epoch 9 - VAL - Batch 300 - Loss = 0.696 | Accuracy = 78.125%
238
+ 2023-01-04 19:50:50,779 - INFO - TRAIN phase
239
+ 2023-01-04 19:50:50,885 - INFO - Epoch 10 - TRAIN - Batch 0 - Loss = 0.618 | Accuracy = 87.5%
240
+ 2023-01-04 19:50:56,980 - INFO - Epoch 10 - TRAIN - Batch 100 - Loss = 0.357 | Accuracy = 87.5%
241
+ 2023-01-04 19:51:02,917 - INFO - Epoch 10 - TRAIN - Batch 200 - Loss = 0.95 | Accuracy = 71.875%
242
+ 2023-01-04 19:51:08,706 - INFO - Epoch 10 - TRAIN - Batch 300 - Loss = 0.41 | Accuracy = 87.5%
243
+ 2023-01-04 19:51:14,627 - INFO - Epoch 10 - TRAIN - Batch 400 - Loss = 0.334 | Accuracy = 90.625%
244
+ 2023-01-04 19:51:20,537 - INFO - Epoch 10 - TRAIN - Batch 500 - Loss = 0.707 | Accuracy = 84.375%
245
+ 2023-01-04 19:51:26,354 - INFO - Epoch 10 - TRAIN - Batch 600 - Loss = 0.599 | Accuracy = 81.25%
246
+ 2023-01-04 19:51:32,134 - INFO - Epoch 10 - TRAIN - Batch 700 - Loss = 0.691 | Accuracy = 90.625%
247
+ 2023-01-04 19:51:38,093 - INFO - Epoch 10 - TRAIN - Batch 800 - Loss = 0.601 | Accuracy = 81.25%
248
+ 2023-01-04 19:51:43,944 - INFO - Epoch 10 - TRAIN - Batch 900 - Loss = 0.623 | Accuracy = 81.25%
249
+ 2023-01-04 19:51:49,707 - INFO - Epoch 10 - TRAIN - Batch 1000 - Loss = 0.548 | Accuracy = 87.5%
250
+ 2023-01-04 19:51:55,506 - INFO - Epoch 10 - TRAIN - Batch 1100 - Loss = 0.723 | Accuracy = 81.25%
251
+ 2023-01-04 19:52:01,336 - INFO - Epoch 10 - TRAIN - Batch 1200 - Loss = 0.414 | Accuracy = 87.5%
252
+ 2023-01-04 19:52:07,155 - INFO - Epoch 10 - TRAIN - Batch 1300 - Loss = 0.289 | Accuracy = 90.625%
253
+ 2023-01-04 19:52:12,905 - INFO - Epoch 10 - TRAIN - Batch 1400 - Loss = 0.899 | Accuracy = 81.25%
254
+ 2023-01-04 19:52:18,736 - INFO - Epoch 10 - TRAIN - Batch 1500 - Loss = 0.601 | Accuracy = 81.25%
255
+ 2023-01-04 19:52:20,566 - INFO - VAL phase
256
+ 2023-01-04 19:52:20,619 - INFO - Epoch 10 - VAL - Batch 0 - Loss = 0.529 | Accuracy = 81.25%
257
+ 2023-01-04 19:52:25,409 - INFO - Epoch 10 - VAL - Batch 100 - Loss = 0.501 | Accuracy = 90.625%
258
+ 2023-01-04 19:52:30,140 - INFO - Epoch 10 - VAL - Batch 200 - Loss = 0.475 | Accuracy = 84.375%
259
+ 2023-01-04 19:52:34,901 - INFO - Epoch 10 - VAL - Batch 300 - Loss = 0.709 | Accuracy = 87.5%
260
+ 2023-01-04 19:52:38,739 - INFO - TRAIN phase
261
+ 2023-01-04 19:52:38,803 - INFO - Epoch 11 - TRAIN - Batch 0 - Loss = 0.694 | Accuracy = 75.0%
262
+ 2023-01-04 19:52:44,632 - INFO - Epoch 11 - TRAIN - Batch 100 - Loss = 0.437 | Accuracy = 90.625%
263
+ 2023-01-04 19:52:50,480 - INFO - Epoch 11 - TRAIN - Batch 200 - Loss = 0.686 | Accuracy = 78.125%
264
+ 2023-01-04 19:52:56,310 - INFO - Epoch 11 - TRAIN - Batch 300 - Loss = 0.339 | Accuracy = 90.625%
265
+ 2023-01-04 19:53:02,246 - INFO - Epoch 11 - TRAIN - Batch 400 - Loss = 0.285 | Accuracy = 90.625%
266
+ 2023-01-04 19:53:08,054 - INFO - Epoch 11 - TRAIN - Batch 500 - Loss = 0.521 | Accuracy = 81.25%
267
+ 2023-01-04 19:53:13,233 - INFO - Epoch 11 - TRAIN - Batch 600 - Loss = 0.77 | Accuracy = 81.25%
268
+ 2023-01-04 19:53:18,358 - INFO - Epoch 11 - TRAIN - Batch 700 - Loss = 0.49 | Accuracy = 84.375%
269
+ 2023-01-04 19:53:23,551 - INFO - Epoch 11 - TRAIN - Batch 800 - Loss = 0.622 | Accuracy = 75.0%
270
+ 2023-01-04 19:53:28,708 - INFO - Epoch 11 - TRAIN - Batch 900 - Loss = 0.764 | Accuracy = 78.125%
271
+ 2023-01-04 19:53:33,935 - INFO - Epoch 11 - TRAIN - Batch 1000 - Loss = 0.505 | Accuracy = 84.375%
272
+ 2023-01-04 19:53:39,104 - INFO - Epoch 11 - TRAIN - Batch 1100 - Loss = 0.761 | Accuracy = 84.375%
273
+ 2023-01-04 19:53:44,373 - INFO - Epoch 11 - TRAIN - Batch 1200 - Loss = 0.484 | Accuracy = 84.375%
274
+ 2023-01-04 19:53:49,531 - INFO - Epoch 11 - TRAIN - Batch 1300 - Loss = 0.403 | Accuracy = 90.625%
275
+ 2023-01-04 19:53:55,216 - INFO - Epoch 11 - TRAIN - Batch 1400 - Loss = 0.495 | Accuracy = 93.75%
276
+ 2023-01-04 19:54:01,069 - INFO - Epoch 11 - TRAIN - Batch 1500 - Loss = 0.208 | Accuracy = 93.75%
277
+ 2023-01-04 19:54:02,783 - INFO - VAL phase
278
+ 2023-01-04 19:54:02,833 - INFO - Epoch 11 - VAL - Batch 0 - Loss = 0.631 | Accuracy = 84.375%
279
+ 2023-01-04 19:54:07,766 - INFO - Epoch 11 - VAL - Batch 100 - Loss = 0.289 | Accuracy = 90.625%
280
+ 2023-01-04 19:54:12,635 - INFO - Epoch 11 - VAL - Batch 200 - Loss = 0.332 | Accuracy = 90.625%
281
+ 2023-01-04 19:54:17,467 - INFO - Epoch 11 - VAL - Batch 300 - Loss = 0.493 | Accuracy = 84.375%
282
+ 2023-01-04 19:54:21,397 - INFO - TRAIN phase
283
+ 2023-01-04 19:54:21,463 - INFO - Epoch 12 - TRAIN - Batch 0 - Loss = 0.319 | Accuracy = 90.625%
284
+ 2023-01-04 19:54:27,223 - INFO - Epoch 12 - TRAIN - Batch 100 - Loss = 0.383 | Accuracy = 93.75%
285
+ 2023-01-04 19:54:33,062 - INFO - Epoch 12 - TRAIN - Batch 200 - Loss = 0.137 | Accuracy = 96.875%
286
+ 2023-01-04 19:54:38,810 - INFO - Epoch 12 - TRAIN - Batch 300 - Loss = 0.349 | Accuracy = 93.75%
287
+ 2023-01-04 19:54:44,666 - INFO - Epoch 12 - TRAIN - Batch 400 - Loss = 0.211 | Accuracy = 93.75%
288
+ 2023-01-04 19:54:50,546 - INFO - Epoch 12 - TRAIN - Batch 500 - Loss = 0.328 | Accuracy = 87.5%
289
+ 2023-01-04 19:54:56,409 - INFO - Epoch 12 - TRAIN - Batch 600 - Loss = 0.162 | Accuracy = 96.875%
290
+ 2023-01-04 19:55:02,289 - INFO - Epoch 12 - TRAIN - Batch 700 - Loss = 0.819 | Accuracy = 81.25%
291
+ 2023-01-04 19:55:08,123 - INFO - Epoch 12 - TRAIN - Batch 800 - Loss = 0.53 | Accuracy = 90.625%
292
+ 2023-01-04 19:55:13,984 - INFO - Epoch 12 - TRAIN - Batch 900 - Loss = 0.781 | Accuracy = 81.25%
293
+ 2023-01-04 19:55:19,892 - INFO - Epoch 12 - TRAIN - Batch 1000 - Loss = 0.797 | Accuracy = 78.125%
294
+ 2023-01-04 19:55:25,760 - INFO - Epoch 12 - TRAIN - Batch 1100 - Loss = 0.525 | Accuracy = 87.5%
295
+ 2023-01-04 19:55:31,283 - INFO - Epoch 12 - TRAIN - Batch 1200 - Loss = 0.379 | Accuracy = 90.625%
296
+ 2023-01-04 19:55:36,479 - INFO - Epoch 12 - TRAIN - Batch 1300 - Loss = 0.711 | Accuracy = 84.375%
297
+ 2023-01-04 19:55:41,686 - INFO - Epoch 12 - TRAIN - Batch 1400 - Loss = 0.222 | Accuracy = 90.625%
298
+ 2023-01-04 19:55:46,951 - INFO - Epoch 12 - TRAIN - Batch 1500 - Loss = 0.486 | Accuracy = 84.375%
299
+ 2023-01-04 19:55:48,440 - INFO - VAL phase
300
+ 2023-01-04 19:55:48,489 - INFO - Epoch 12 - VAL - Batch 0 - Loss = 0.244 | Accuracy = 90.625%
301
+ 2023-01-04 19:55:53,070 - INFO - Epoch 12 - VAL - Batch 100 - Loss = 0.395 | Accuracy = 87.5%
302
+ 2023-01-04 19:55:57,642 - INFO - Epoch 12 - VAL - Batch 200 - Loss = 0.295 | Accuracy = 90.625%
303
+ 2023-01-04 19:56:02,237 - INFO - Epoch 12 - VAL - Batch 300 - Loss = 0.491 | Accuracy = 90.625%
304
+ 2023-01-04 19:56:05,952 - INFO - TRAIN phase
305
+ 2023-01-04 19:56:06,016 - INFO - Epoch 13 - TRAIN - Batch 0 - Loss = 0.49 | Accuracy = 78.125%
306
+ 2023-01-04 19:56:11,270 - INFO - Epoch 13 - TRAIN - Batch 100 - Loss = 0.672 | Accuracy = 71.875%
307
+ 2023-01-04 19:56:16,891 - INFO - Epoch 13 - TRAIN - Batch 200 - Loss = 0.256 | Accuracy = 93.75%
308
+ 2023-01-04 19:56:21,887 - INFO - Epoch 13 - TRAIN - Batch 300 - Loss = 0.272 | Accuracy = 93.75%
309
+ 2023-01-04 19:56:26,991 - INFO - Epoch 13 - TRAIN - Batch 400 - Loss = 0.565 | Accuracy = 81.25%
310
+ 2023-01-04 19:56:32,309 - INFO - Epoch 13 - TRAIN - Batch 500 - Loss = 0.173 | Accuracy = 93.75%
311
+ 2023-01-04 19:56:37,470 - INFO - Epoch 13 - TRAIN - Batch 600 - Loss = 0.472 | Accuracy = 84.375%
312
+ 2023-01-04 19:56:42,590 - INFO - Epoch 13 - TRAIN - Batch 700 - Loss = 0.457 | Accuracy = 87.5%
313
+ 2023-01-04 19:56:47,774 - INFO - Epoch 13 - TRAIN - Batch 800 - Loss = 0.417 | Accuracy = 87.5%
314
+ 2023-01-04 19:56:52,944 - INFO - Epoch 13 - TRAIN - Batch 900 - Loss = 0.412 | Accuracy = 90.625%
315
+ 2023-01-04 19:56:58,158 - INFO - Epoch 13 - TRAIN - Batch 1000 - Loss = 0.541 | Accuracy = 81.25%
316
+ 2023-01-04 19:57:03,609 - INFO - Epoch 13 - TRAIN - Batch 1100 - Loss = 0.242 | Accuracy = 87.5%
317
+ 2023-01-04 19:57:08,614 - INFO - Epoch 13 - TRAIN - Batch 1200 - Loss = 0.198 | Accuracy = 93.75%
318
+ 2023-01-04 19:57:13,357 - INFO - Epoch 13 - TRAIN - Batch 1300 - Loss = 0.642 | Accuracy = 78.125%
319
+ 2023-01-04 19:57:18,186 - INFO - Epoch 13 - TRAIN - Batch 1400 - Loss = 0.274 | Accuracy = 90.625%
320
+ 2023-01-04 19:57:22,956 - INFO - Epoch 13 - TRAIN - Batch 1500 - Loss = 0.242 | Accuracy = 90.625%
321
+ 2023-01-04 19:57:24,375 - INFO - VAL phase
322
+ 2023-01-04 19:57:24,426 - INFO - Epoch 13 - VAL - Batch 0 - Loss = 0.626 | Accuracy = 87.5%
323
+ 2023-01-04 19:57:28,979 - INFO - Epoch 13 - VAL - Batch 100 - Loss = 0.299 | Accuracy = 84.375%
324
+ 2023-01-04 19:57:33,440 - INFO - Epoch 13 - VAL - Batch 200 - Loss = 0.342 | Accuracy = 90.625%
325
+ 2023-01-04 19:57:37,897 - INFO - Epoch 13 - VAL - Batch 300 - Loss = 0.63 | Accuracy = 87.5%
326
+ 2023-01-04 19:57:41,507 - INFO - TRAIN phase
327
+ 2023-01-04 19:57:41,572 - INFO - Epoch 14 - TRAIN - Batch 0 - Loss = 0.276 | Accuracy = 87.5%
328
+ 2023-01-04 19:57:46,452 - INFO - Epoch 14 - TRAIN - Batch 100 - Loss = 0.595 | Accuracy = 81.25%
329
+ 2023-01-04 19:57:51,467 - INFO - Epoch 14 - TRAIN - Batch 200 - Loss = 0.266 | Accuracy = 93.75%
330
+ 2023-01-04 19:57:57,008 - INFO - Epoch 14 - TRAIN - Batch 300 - Loss = 0.537 | Accuracy = 84.375%
331
+ 2023-01-04 19:58:02,216 - INFO - Epoch 14 - TRAIN - Batch 400 - Loss = 0.243 | Accuracy = 93.75%
332
+ 2023-01-04 19:58:07,375 - INFO - Epoch 14 - TRAIN - Batch 500 - Loss = 0.216 | Accuracy = 93.75%
333
+ 2023-01-04 19:58:12,696 - INFO - Epoch 14 - TRAIN - Batch 600 - Loss = 0.485 | Accuracy = 81.25%
334
+ 2023-01-04 19:58:17,917 - INFO - Epoch 14 - TRAIN - Batch 700 - Loss = 0.46 | Accuracy = 90.625%
335
+ 2023-01-04 19:58:23,156 - INFO - Epoch 14 - TRAIN - Batch 800 - Loss = 1.091 | Accuracy = 81.25%
336
+ 2023-01-04 19:58:28,368 - INFO - Epoch 14 - TRAIN - Batch 900 - Loss = 0.414 | Accuracy = 90.625%
337
+ 2023-01-04 19:58:33,620 - INFO - Epoch 14 - TRAIN - Batch 1000 - Loss = 0.866 | Accuracy = 81.25%
338
+ 2023-01-04 19:58:38,748 - INFO - Epoch 14 - TRAIN - Batch 1100 - Loss = 0.555 | Accuracy = 81.25%
339
+ 2023-01-04 19:58:44,170 - INFO - Epoch 14 - TRAIN - Batch 1200 - Loss = 0.877 | Accuracy = 78.125%
340
+ 2023-01-04 19:58:50,045 - INFO - Epoch 14 - TRAIN - Batch 1300 - Loss = 0.565 | Accuracy = 81.25%
341
+ 2023-01-04 19:58:55,828 - INFO - Epoch 14 - TRAIN - Batch 1400 - Loss = 0.7 | Accuracy = 81.25%
342
+ 2023-01-04 19:59:01,728 - INFO - Epoch 14 - TRAIN - Batch 1500 - Loss = 0.096 | Accuracy = 100.0%
343
+ 2023-01-04 19:59:03,419 - INFO - VAL phase
344
+ 2023-01-04 19:59:03,468 - INFO - Epoch 14 - VAL - Batch 0 - Loss = 0.115 | Accuracy = 96.875%
345
+ 2023-01-04 19:59:08,228 - INFO - Epoch 14 - VAL - Batch 100 - Loss = 0.456 | Accuracy = 90.625%
346
+ 2023-01-04 19:59:13,053 - INFO - Epoch 14 - VAL - Batch 200 - Loss = 0.346 | Accuracy = 90.625%
347
+ 2023-01-04 19:59:17,885 - INFO - Epoch 14 - VAL - Batch 300 - Loss = 0.38 | Accuracy = 90.625%
348
+ 2023-01-04 19:59:21,782 - INFO - TRAIN phase
349
+ 2023-01-04 19:59:21,855 - INFO - Epoch 15 - TRAIN - Batch 0 - Loss = 0.481 | Accuracy = 75.0%
350
+ 2023-01-04 19:59:27,637 - INFO - Epoch 15 - TRAIN - Batch 100 - Loss = 0.514 | Accuracy = 84.375%
351
+ 2023-01-04 19:59:33,492 - INFO - Epoch 15 - TRAIN - Batch 200 - Loss = 0.676 | Accuracy = 87.5%
352
+ 2023-01-04 19:59:39,256 - INFO - Epoch 15 - TRAIN - Batch 300 - Loss = 0.43 | Accuracy = 87.5%
353
+ 2023-01-04 19:59:45,100 - INFO - Epoch 15 - TRAIN - Batch 400 - Loss = 0.594 | Accuracy = 81.25%
354
+ 2023-01-04 19:59:50,930 - INFO - Epoch 15 - TRAIN - Batch 500 - Loss = 0.444 | Accuracy = 84.375%
355
+ 2023-01-04 19:59:56,714 - INFO - Epoch 15 - TRAIN - Batch 600 - Loss = 0.325 | Accuracy = 93.75%
356
+ 2023-01-04 20:00:02,733 - INFO - Epoch 15 - TRAIN - Batch 700 - Loss = 0.675 | Accuracy = 78.125%
357
+ 2023-01-04 20:00:08,566 - INFO - Epoch 15 - TRAIN - Batch 800 - Loss = 0.363 | Accuracy = 81.25%
358
+ 2023-01-04 20:00:14,271 - INFO - Epoch 15 - TRAIN - Batch 900 - Loss = 0.609 | Accuracy = 81.25%
359
+ 2023-01-04 20:00:20,193 - INFO - Epoch 15 - TRAIN - Batch 1000 - Loss = 0.162 | Accuracy = 96.875%
360
+ 2023-01-04 20:00:26,034 - INFO - Epoch 15 - TRAIN - Batch 1100 - Loss = 0.391 | Accuracy = 81.25%
361
+ 2023-01-04 20:00:31,731 - INFO - Epoch 15 - TRAIN - Batch 1200 - Loss = 0.504 | Accuracy = 81.25%
362
+ 2023-01-04 20:00:37,557 - INFO - Epoch 15 - TRAIN - Batch 1300 - Loss = 0.559 | Accuracy = 84.375%
363
+ 2023-01-04 20:00:43,383 - INFO - Epoch 15 - TRAIN - Batch 1400 - Loss = 0.271 | Accuracy = 93.75%
364
+ 2023-01-04 20:00:49,144 - INFO - Epoch 15 - TRAIN - Batch 1500 - Loss = 0.445 | Accuracy = 87.5%
365
+ 2023-01-04 20:00:50,829 - INFO - VAL phase
366
+ 2023-01-04 20:00:50,888 - INFO - Epoch 15 - VAL - Batch 0 - Loss = 0.524 | Accuracy = 87.5%
367
+ 2023-01-04 20:00:55,715 - INFO - Epoch 15 - VAL - Batch 100 - Loss = 0.257 | Accuracy = 93.75%
368
+ 2023-01-04 20:01:00,439 - INFO - Epoch 15 - VAL - Batch 200 - Loss = 0.468 | Accuracy = 84.375%
369
+ 2023-01-04 20:01:05,236 - INFO - Epoch 15 - VAL - Batch 300 - Loss = 0.453 | Accuracy = 87.5%
370
+ 2023-01-04 20:01:09,713 - INFO - TRAIN phase
371
+ 2023-01-04 20:01:09,843 - INFO - Epoch 16 - TRAIN - Batch 0 - Loss = 0.647 | Accuracy = 81.25%
372
+ 2023-01-04 20:01:14,705 - INFO - Epoch 16 - TRAIN - Batch 100 - Loss = 0.27 | Accuracy = 90.625%
373
+ 2023-01-04 20:01:19,542 - INFO - Epoch 16 - TRAIN - Batch 200 - Loss = 0.284 | Accuracy = 93.75%
374
+ 2023-01-04 20:01:24,304 - INFO - Epoch 16 - TRAIN - Batch 300 - Loss = 0.559 | Accuracy = 81.25%
375
+ 2023-01-04 20:01:29,088 - INFO - Epoch 16 - TRAIN - Batch 400 - Loss = 0.272 | Accuracy = 90.625%
376
+ 2023-01-04 20:01:33,888 - INFO - Epoch 16 - TRAIN - Batch 500 - Loss = 0.523 | Accuracy = 81.25%
377
+ 2023-01-04 20:01:38,654 - INFO - Epoch 16 - TRAIN - Batch 600 - Loss = 0.257 | Accuracy = 87.5%
378
+ 2023-01-04 20:01:43,412 - INFO - Epoch 16 - TRAIN - Batch 700 - Loss = 0.48 | Accuracy = 90.625%
379
+ 2023-01-04 20:01:48,229 - INFO - Epoch 16 - TRAIN - Batch 800 - Loss = 0.781 | Accuracy = 81.25%
380
+ 2023-01-04 20:01:53,042 - INFO - Epoch 16 - TRAIN - Batch 900 - Loss = 0.477 | Accuracy = 90.625%
381
+ 2023-01-04 20:01:58,611 - INFO - Epoch 16 - TRAIN - Batch 1000 - Loss = 0.226 | Accuracy = 93.75%
382
+ 2023-01-04 20:02:03,991 - INFO - Epoch 16 - TRAIN - Batch 1100 - Loss = 0.552 | Accuracy = 87.5%
383
+ 2023-01-04 20:02:09,276 - INFO - Epoch 16 - TRAIN - Batch 1200 - Loss = 0.396 | Accuracy = 84.375%
384
+ 2023-01-04 20:02:14,465 - INFO - Epoch 16 - TRAIN - Batch 1300 - Loss = 0.468 | Accuracy = 90.625%
385
+ 2023-01-04 20:02:19,582 - INFO - Epoch 16 - TRAIN - Batch 1400 - Loss = 0.228 | Accuracy = 93.75%
386
+ 2023-01-04 20:02:24,726 - INFO - Epoch 16 - TRAIN - Batch 1500 - Loss = 0.603 | Accuracy = 78.125%
387
+ 2023-01-04 20:02:26,288 - INFO - VAL phase
388
+ 2023-01-04 20:02:26,347 - INFO - Epoch 16 - VAL - Batch 0 - Loss = 0.319 | Accuracy = 90.625%
389
+ 2023-01-04 20:02:30,890 - INFO - Epoch 16 - VAL - Batch 100 - Loss = 0.38 | Accuracy = 90.625%
390
+ 2023-01-04 20:02:35,404 - INFO - Epoch 16 - VAL - Batch 200 - Loss = 0.56 | Accuracy = 81.25%
391
+ 2023-01-04 20:02:39,993 - INFO - Epoch 16 - VAL - Batch 300 - Loss = 0.331 | Accuracy = 93.75%
392
+ 2023-01-04 20:02:43,790 - INFO - TRAIN phase
393
+ 2023-01-04 20:02:43,868 - INFO - Epoch 17 - TRAIN - Batch 0 - Loss = 0.469 | Accuracy = 81.25%
394
+ 2023-01-04 20:02:52,791 - INFO - Epoch 17 - TRAIN - Batch 100 - Loss = 0.43 | Accuracy = 84.375%
395
+ 2023-01-04 20:02:57,938 - INFO - Epoch 17 - TRAIN - Batch 200 - Loss = 0.209 | Accuracy = 93.75%
396
+ 2023-01-04 20:03:03,214 - INFO - Epoch 17 - TRAIN - Batch 300 - Loss = 0.164 | Accuracy = 90.625%
397
+ 2023-01-04 20:03:08,500 - INFO - Epoch 17 - TRAIN - Batch 400 - Loss = 0.548 | Accuracy = 81.25%
398
+ 2023-01-04 20:03:13,742 - INFO - Epoch 17 - TRAIN - Batch 500 - Loss = 0.379 | Accuracy = 87.5%
399
+ 2023-01-04 20:03:19,061 - INFO - Epoch 17 - TRAIN - Batch 600 - Loss = 0.162 | Accuracy = 96.875%
400
+ 2023-01-04 20:03:24,411 - INFO - Epoch 17 - TRAIN - Batch 700 - Loss = 0.497 | Accuracy = 81.25%
401
+ 2023-01-04 20:03:29,690 - INFO - Epoch 17 - TRAIN - Batch 800 - Loss = 0.422 | Accuracy = 90.625%
402
+ 2023-01-04 20:03:35,502 - INFO - Epoch 17 - TRAIN - Batch 900 - Loss = 0.401 | Accuracy = 93.75%
403
+ 2023-01-04 20:03:41,311 - INFO - Epoch 17 - TRAIN - Batch 1000 - Loss = 0.458 | Accuracy = 81.25%
404
+ 2023-01-04 20:03:47,166 - INFO - Epoch 17 - TRAIN - Batch 1100 - Loss = 0.398 | Accuracy = 87.5%
405
+ 2023-01-04 20:03:53,134 - INFO - Epoch 17 - TRAIN - Batch 1200 - Loss = 0.709 | Accuracy = 81.25%
406
+ 2023-01-04 20:03:59,007 - INFO - Epoch 17 - TRAIN - Batch 1300 - Loss = 0.39 | Accuracy = 81.25%
407
+ 2023-01-04 20:04:05,006 - INFO - Epoch 17 - TRAIN - Batch 1400 - Loss = 0.469 | Accuracy = 81.25%
408
+ 2023-01-04 20:04:10,887 - INFO - Epoch 17 - TRAIN - Batch 1500 - Loss = 0.384 | Accuracy = 90.625%
409
+ 2023-01-04 20:04:12,539 - INFO - VAL phase
410
+ 2023-01-04 20:04:12,587 - INFO - Epoch 17 - VAL - Batch 0 - Loss = 0.24 | Accuracy = 90.625%
411
+ 2023-01-04 20:04:17,400 - INFO - Epoch 17 - VAL - Batch 100 - Loss = 0.52 | Accuracy = 87.5%
412
+ 2023-01-04 20:04:22,539 - INFO - Epoch 17 - VAL - Batch 200 - Loss = 0.394 | Accuracy = 87.5%
413
+ 2023-01-04 20:04:28,169 - INFO - Epoch 17 - VAL - Batch 300 - Loss = 0.485 | Accuracy = 84.375%
414
+ 2023-01-04 20:04:32,847 - INFO - TRAIN phase
415
+ 2023-01-04 20:04:32,954 - INFO - Epoch 18 - TRAIN - Batch 0 - Loss = 0.401 | Accuracy = 84.375%
416
+ 2023-01-04 20:04:37,943 - INFO - Epoch 18 - TRAIN - Batch 100 - Loss = 0.635 | Accuracy = 81.25%
417
+ 2023-01-04 20:04:42,886 - INFO - Epoch 18 - TRAIN - Batch 200 - Loss = 0.365 | Accuracy = 87.5%
418
+ 2023-01-04 20:04:47,852 - INFO - Epoch 18 - TRAIN - Batch 300 - Loss = 0.74 | Accuracy = 87.5%
419
+ 2023-01-04 20:04:52,723 - INFO - Epoch 18 - TRAIN - Batch 400 - Loss = 0.401 | Accuracy = 87.5%
420
+ 2023-01-04 20:04:57,611 - INFO - Epoch 18 - TRAIN - Batch 500 - Loss = 0.603 | Accuracy = 84.375%
421
+ 2023-01-04 20:05:02,429 - INFO - Epoch 18 - TRAIN - Batch 600 - Loss = 0.553 | Accuracy = 81.25%
422
+ 2023-01-04 20:05:07,199 - INFO - Epoch 18 - TRAIN - Batch 700 - Loss = 1.159 | Accuracy = 71.875%
423
+ 2023-01-04 20:05:12,122 - INFO - Epoch 18 - TRAIN - Batch 800 - Loss = 0.466 | Accuracy = 90.625%
424
+ 2023-01-04 20:05:17,084 - INFO - Epoch 18 - TRAIN - Batch 900 - Loss = 0.466 | Accuracy = 81.25%
425
+ 2023-01-04 20:05:22,939 - INFO - Epoch 18 - TRAIN - Batch 1000 - Loss = 0.446 | Accuracy = 84.375%
426
+ 2023-01-04 20:05:28,820 - INFO - Epoch 18 - TRAIN - Batch 1100 - Loss = 0.378 | Accuracy = 84.375%
427
+ 2023-01-04 20:05:34,715 - INFO - Epoch 18 - TRAIN - Batch 1200 - Loss = 0.302 | Accuracy = 90.625%
428
+ 2023-01-04 20:05:40,577 - INFO - Epoch 18 - TRAIN - Batch 1300 - Loss = 0.678 | Accuracy = 84.375%
429
+ 2023-01-04 20:05:46,433 - INFO - Epoch 18 - TRAIN - Batch 1400 - Loss = 0.358 | Accuracy = 90.625%
430
+ 2023-01-04 20:05:52,222 - INFO - Epoch 18 - TRAIN - Batch 1500 - Loss = 0.451 | Accuracy = 87.5%
431
+ 2023-01-04 20:05:53,916 - INFO - VAL phase
432
+ 2023-01-04 20:05:53,972 - INFO - Epoch 18 - VAL - Batch 0 - Loss = 0.27 | Accuracy = 93.75%
433
+ 2023-01-04 20:05:58,827 - INFO - Epoch 18 - VAL - Batch 100 - Loss = 0.321 | Accuracy = 87.5%
434
+ 2023-01-04 20:06:03,684 - INFO - Epoch 18 - VAL - Batch 200 - Loss = 0.253 | Accuracy = 93.75%
435
+ 2023-01-04 20:06:08,523 - INFO - Epoch 18 - VAL - Batch 300 - Loss = 0.248 | Accuracy = 90.625%
436
+ 2023-01-04 20:06:13,130 - INFO - TRAIN phase
437
+ 2023-01-04 20:06:13,263 - INFO - Epoch 19 - TRAIN - Batch 0 - Loss = 0.452 | Accuracy = 87.5%
438
+ 2023-01-04 20:06:18,154 - INFO - Epoch 19 - TRAIN - Batch 100 - Loss = 0.25 | Accuracy = 93.75%
439
+ 2023-01-04 20:06:23,011 - INFO - Epoch 19 - TRAIN - Batch 200 - Loss = 0.815 | Accuracy = 78.125%
440
+ 2023-01-04 20:06:27,839 - INFO - Epoch 19 - TRAIN - Batch 300 - Loss = 0.184 | Accuracy = 93.75%
441
+ 2023-01-04 20:06:32,665 - INFO - Epoch 19 - TRAIN - Batch 400 - Loss = 0.385 | Accuracy = 87.5%
442
+ 2023-01-04 20:06:37,446 - INFO - Epoch 19 - TRAIN - Batch 500 - Loss = 0.373 | Accuracy = 81.25%
443
+ 2023-01-04 20:06:42,272 - INFO - Epoch 19 - TRAIN - Batch 600 - Loss = 0.558 | Accuracy = 81.25%
444
+ 2023-01-04 20:06:47,130 - INFO - Epoch 19 - TRAIN - Batch 700 - Loss = 0.497 | Accuracy = 81.25%
445
+ 2023-01-04 20:06:51,881 - INFO - Epoch 19 - TRAIN - Batch 800 - Loss = 0.325 | Accuracy = 90.625%
446
+ 2023-01-04 20:06:56,668 - INFO - Epoch 19 - TRAIN - Batch 900 - Loss = 0.323 | Accuracy = 81.25%
447
+ 2023-01-04 20:07:02,369 - INFO - Epoch 19 - TRAIN - Batch 1000 - Loss = 0.54 | Accuracy = 84.375%
448
+ 2023-01-04 20:07:07,608 - INFO - Epoch 19 - TRAIN - Batch 1100 - Loss = 0.697 | Accuracy = 78.125%
449
+ 2023-01-04 20:07:12,749 - INFO - Epoch 19 - TRAIN - Batch 1200 - Loss = 0.474 | Accuracy = 84.375%
450
+ 2023-01-04 20:07:17,961 - INFO - Epoch 19 - TRAIN - Batch 1300 - Loss = 0.292 | Accuracy = 93.75%
451
+ 2023-01-04 20:07:23,233 - INFO - Epoch 19 - TRAIN - Batch 1400 - Loss = 0.365 | Accuracy = 87.5%
452
+ 2023-01-04 20:07:28,526 - INFO - Epoch 19 - TRAIN - Batch 1500 - Loss = 0.366 | Accuracy = 90.625%
453
+ 2023-01-04 20:07:30,055 - INFO - VAL phase
454
+ 2023-01-04 20:07:30,109 - INFO - Epoch 19 - VAL - Batch 0 - Loss = 0.706 | Accuracy = 84.375%
455
+ 2023-01-04 20:07:34,576 - INFO - Epoch 19 - VAL - Batch 100 - Loss = 0.703 | Accuracy = 78.125%
456
+ 2023-01-04 20:07:39,249 - INFO - Epoch 19 - VAL - Batch 200 - Loss = 0.313 | Accuracy = 87.5%
457
+ 2023-01-04 20:07:43,725 - INFO - Epoch 19 - VAL - Batch 300 - Loss = 0.234 | Accuracy = 90.625%
458
+ 2023-01-04 20:07:47,547 - INFO - TRAIN phase
459
+ 2023-01-04 20:07:47,620 - INFO - Epoch 20 - TRAIN - Batch 0 - Loss = 0.203 | Accuracy = 93.75%
460
+ 2023-01-04 20:07:53,698 - INFO - Epoch 20 - TRAIN - Batch 100 - Loss = 0.639 | Accuracy = 90.625%
461
+ 2023-01-04 20:07:59,712 - INFO - Epoch 20 - TRAIN - Batch 200 - Loss = 0.456 | Accuracy = 84.375%
462
+ 2023-01-04 20:08:05,597 - INFO - Epoch 20 - TRAIN - Batch 300 - Loss = 0.464 | Accuracy = 87.5%
463
+ 2023-01-04 20:08:11,332 - INFO - Epoch 20 - TRAIN - Batch 400 - Loss = 0.544 | Accuracy = 90.625%
464
+ 2023-01-04 20:08:17,232 - INFO - Epoch 20 - TRAIN - Batch 500 - Loss = 0.369 | Accuracy = 90.625%
465
+ 2023-01-04 20:08:23,051 - INFO - Epoch 20 - TRAIN - Batch 600 - Loss = 0.342 | Accuracy = 93.75%
466
+ 2023-01-04 20:08:28,840 - INFO - Epoch 20 - TRAIN - Batch 700 - Loss = 0.406 | Accuracy = 90.625%
467
+ 2023-01-04 20:08:34,646 - INFO - Epoch 20 - TRAIN - Batch 800 - Loss = 0.668 | Accuracy = 75.0%
468
+ 2023-01-04 20:08:40,287 - INFO - Epoch 20 - TRAIN - Batch 900 - Loss = 0.197 | Accuracy = 93.75%
469
+ 2023-01-04 20:08:45,511 - INFO - Epoch 20 - TRAIN - Batch 1000 - Loss = 0.743 | Accuracy = 90.625%
470
+ 2023-01-04 20:08:50,809 - INFO - Epoch 20 - TRAIN - Batch 1100 - Loss = 0.584 | Accuracy = 84.375%
471
+ 2023-01-04 20:08:56,061 - INFO - Epoch 20 - TRAIN - Batch 1200 - Loss = 0.377 | Accuracy = 90.625%
472
+ 2023-01-04 20:09:01,321 - INFO - Epoch 20 - TRAIN - Batch 1300 - Loss = 0.806 | Accuracy = 81.25%
473
+ 2023-01-04 20:09:06,468 - INFO - Epoch 20 - TRAIN - Batch 1400 - Loss = 0.4 | Accuracy = 90.625%
474
+ 2023-01-04 20:09:11,688 - INFO - Epoch 20 - TRAIN - Batch 1500 - Loss = 0.595 | Accuracy = 90.625%
475
+ 2023-01-04 20:09:13,206 - INFO - VAL phase
476
+ 2023-01-04 20:09:13,264 - INFO - Epoch 20 - VAL - Batch 0 - Loss = 0.178 | Accuracy = 93.75%
477
+ 2023-01-04 20:09:17,940 - INFO - Epoch 20 - VAL - Batch 100 - Loss = 0.367 | Accuracy = 84.375%
478
+ 2023-01-04 20:09:22,552 - INFO - Epoch 20 - VAL - Batch 200 - Loss = 0.381 | Accuracy = 93.75%
479
+ 2023-01-04 20:09:27,541 - INFO - Epoch 20 - VAL - Batch 300 - Loss = 0.161 | Accuracy = 93.75%
480
+ 2023-01-04 20:09:32,136 - INFO - TRAIN phase
481
+ 2023-01-04 20:09:32,239 - INFO - Epoch 21 - TRAIN - Batch 0 - Loss = 0.355 | Accuracy = 87.5%
482
+ 2023-01-04 20:09:37,698 - INFO - Epoch 21 - TRAIN - Batch 100 - Loss = 0.213 | Accuracy = 96.875%
483
+ 2023-01-04 20:09:42,884 - INFO - Epoch 21 - TRAIN - Batch 200 - Loss = 0.296 | Accuracy = 90.625%
484
+ 2023-01-04 20:09:47,974 - INFO - Epoch 21 - TRAIN - Batch 300 - Loss = 0.928 | Accuracy = 81.25%
485
+ 2023-01-04 20:09:53,131 - INFO - Epoch 21 - TRAIN - Batch 400 - Loss = 0.189 | Accuracy = 96.875%
486
+ 2023-01-04 20:09:58,299 - INFO - Epoch 21 - TRAIN - Batch 500 - Loss = 0.101 | Accuracy = 96.875%
487
+ 2023-01-04 20:10:03,571 - INFO - Epoch 21 - TRAIN - Batch 600 - Loss = 0.177 | Accuracy = 93.75%
488
+ 2023-01-04 20:10:08,761 - INFO - Epoch 21 - TRAIN - Batch 700 - Loss = 0.406 | Accuracy = 84.375%
489
+ 2023-01-04 20:10:13,980 - INFO - Epoch 21 - TRAIN - Batch 800 - Loss = 0.894 | Accuracy = 71.875%
490
+ 2023-01-04 20:10:19,843 - INFO - Epoch 21 - TRAIN - Batch 900 - Loss = 0.558 | Accuracy = 84.375%
491
+ 2023-01-04 20:10:25,131 - INFO - Epoch 21 - TRAIN - Batch 1000 - Loss = 0.436 | Accuracy = 90.625%
492
+ 2023-01-04 20:10:30,165 - INFO - Epoch 21 - TRAIN - Batch 1100 - Loss = 0.422 | Accuracy = 87.5%
493
+ 2023-01-04 20:10:35,343 - INFO - Epoch 21 - TRAIN - Batch 1200 - Loss = 0.347 | Accuracy = 84.375%
494
+ 2023-01-04 20:10:40,578 - INFO - Epoch 21 - TRAIN - Batch 1300 - Loss = 0.209 | Accuracy = 96.875%
495
+ 2023-01-04 20:10:45,788 - INFO - Epoch 21 - TRAIN - Batch 1400 - Loss = 0.527 | Accuracy = 81.25%
496
+ 2023-01-04 20:10:51,011 - INFO - Epoch 21 - TRAIN - Batch 1500 - Loss = 0.756 | Accuracy = 75.0%
497
+ 2023-01-04 20:10:52,545 - INFO - VAL phase
498
+ 2023-01-04 20:10:52,599 - INFO - Epoch 21 - VAL - Batch 0 - Loss = 0.293 | Accuracy = 93.75%
499
+ 2023-01-04 20:10:57,244 - INFO - Epoch 21 - VAL - Batch 100 - Loss = 0.513 | Accuracy = 87.5%
500
+ 2023-01-04 20:11:01,877 - INFO - Epoch 21 - VAL - Batch 200 - Loss = 0.261 | Accuracy = 87.5%
501
+ 2023-01-04 20:11:06,531 - INFO - Epoch 21 - VAL - Batch 300 - Loss = 0.153 | Accuracy = 93.75%
502
+ 2023-01-04 20:11:10,844 - INFO - TRAIN phase
503
+ 2023-01-04 20:11:10,926 - INFO - Epoch 22 - TRAIN - Batch 0 - Loss = 0.218 | Accuracy = 93.75%
504
+ 2023-01-04 20:11:15,837 - INFO - Epoch 22 - TRAIN - Batch 100 - Loss = 0.336 | Accuracy = 90.625%
505
+ 2023-01-04 20:11:20,704 - INFO - Epoch 22 - TRAIN - Batch 200 - Loss = 0.328 | Accuracy = 87.5%
506
+ 2023-01-04 20:11:25,499 - INFO - Epoch 22 - TRAIN - Batch 300 - Loss = 0.706 | Accuracy = 84.375%
507
+ 2023-01-04 20:11:30,244 - INFO - Epoch 22 - TRAIN - Batch 400 - Loss = 0.278 | Accuracy = 96.875%
508
+ 2023-01-04 20:11:35,051 - INFO - Epoch 22 - TRAIN - Batch 500 - Loss = 1.133 | Accuracy = 81.25%
509
+ 2023-01-04 20:11:39,757 - INFO - Epoch 22 - TRAIN - Batch 600 - Loss = 0.512 | Accuracy = 84.375%
510
+ 2023-01-04 20:11:44,601 - INFO - Epoch 22 - TRAIN - Batch 700 - Loss = 0.163 | Accuracy = 93.75%
511
+ 2023-01-04 20:11:49,467 - INFO - Epoch 22 - TRAIN - Batch 800 - Loss = 0.202 | Accuracy = 90.625%
512
+ 2023-01-04 20:11:54,315 - INFO - Epoch 22 - TRAIN - Batch 900 - Loss = 0.144 | Accuracy = 96.875%
513
+ 2023-01-04 20:11:59,395 - INFO - Epoch 22 - TRAIN - Batch 1000 - Loss = 0.139 | Accuracy = 96.875%
514
+ 2023-01-04 20:12:05,371 - INFO - Epoch 22 - TRAIN - Batch 1100 - Loss = 0.327 | Accuracy = 90.625%
515
+ 2023-01-04 20:12:11,206 - INFO - Epoch 22 - TRAIN - Batch 1200 - Loss = 0.458 | Accuracy = 84.375%
516
+ 2023-01-04 20:12:16,985 - INFO - Epoch 22 - TRAIN - Batch 1300 - Loss = 0.28 | Accuracy = 90.625%
517
+ 2023-01-04 20:12:22,808 - INFO - Epoch 22 - TRAIN - Batch 1400 - Loss = 0.269 | Accuracy = 87.5%
518
+ 2023-01-04 20:12:28,598 - INFO - Epoch 22 - TRAIN - Batch 1500 - Loss = 0.465 | Accuracy = 87.5%
519
+ 2023-01-04 20:12:30,291 - INFO - VAL phase
520
+ 2023-01-04 20:12:30,343 - INFO - Epoch 22 - VAL - Batch 0 - Loss = 0.333 | Accuracy = 90.625%
521
+ 2023-01-04 20:12:35,113 - INFO - Epoch 22 - VAL - Batch 100 - Loss = 0.218 | Accuracy = 90.625%
522
+ 2023-01-04 20:12:39,961 - INFO - Epoch 22 - VAL - Batch 200 - Loss = 0.375 | Accuracy = 90.625%
523
+ 2023-01-04 20:12:44,757 - INFO - Epoch 22 - VAL - Batch 300 - Loss = 0.147 | Accuracy = 93.75%
524
+ 2023-01-04 20:12:48,676 - INFO - TRAIN phase
525
+ 2023-01-04 20:12:48,768 - INFO - Epoch 23 - TRAIN - Batch 0 - Loss = 0.451 | Accuracy = 84.375%
526
+ 2023-01-04 20:12:53,830 - INFO - Epoch 23 - TRAIN - Batch 100 - Loss = 0.586 | Accuracy = 84.375%
527
+ 2023-01-04 20:12:58,637 - INFO - Epoch 23 - TRAIN - Batch 200 - Loss = 0.448 | Accuracy = 87.5%
528
+ 2023-01-04 20:13:03,474 - INFO - Epoch 23 - TRAIN - Batch 300 - Loss = 0.569 | Accuracy = 78.125%
529
+ 2023-01-04 20:13:08,336 - INFO - Epoch 23 - TRAIN - Batch 400 - Loss = 1.027 | Accuracy = 81.25%
530
+ 2023-01-04 20:13:13,149 - INFO - Epoch 23 - TRAIN - Batch 500 - Loss = 0.358 | Accuracy = 90.625%
531
+ 2023-01-04 20:13:17,981 - INFO - Epoch 23 - TRAIN - Batch 600 - Loss = 0.311 | Accuracy = 90.625%
532
+ 2023-01-04 20:13:22,826 - INFO - Epoch 23 - TRAIN - Batch 700 - Loss = 0.465 | Accuracy = 87.5%
533
+ 2023-01-04 20:13:27,643 - INFO - Epoch 23 - TRAIN - Batch 800 - Loss = 0.548 | Accuracy = 81.25%
534
+ 2023-01-04 20:13:32,448 - INFO - Epoch 23 - TRAIN - Batch 900 - Loss = 0.29 | Accuracy = 90.625%
535
+ 2023-01-04 20:13:37,354 - INFO - Epoch 23 - TRAIN - Batch 1000 - Loss = 0.373 | Accuracy = 81.25%
536
+ 2023-01-04 20:13:43,251 - INFO - Epoch 23 - TRAIN - Batch 1100 - Loss = 0.712 | Accuracy = 81.25%
537
+ 2023-01-04 20:13:49,094 - INFO - Epoch 23 - TRAIN - Batch 1200 - Loss = 0.178 | Accuracy = 96.875%
538
+ 2023-01-04 20:13:54,883 - INFO - Epoch 23 - TRAIN - Batch 1300 - Loss = 0.286 | Accuracy = 90.625%
539
+ 2023-01-04 20:14:00,737 - INFO - Epoch 23 - TRAIN - Batch 1400 - Loss = 0.395 | Accuracy = 87.5%
540
+ 2023-01-04 20:14:06,571 - INFO - Epoch 23 - TRAIN - Batch 1500 - Loss = 0.139 | Accuracy = 96.875%
541
+ 2023-01-04 20:14:08,272 - INFO - VAL phase
542
+ 2023-01-04 20:14:08,326 - INFO - Epoch 23 - VAL - Batch 0 - Loss = 0.755 | Accuracy = 81.25%
543
+ 2023-01-04 20:14:12,987 - INFO - Epoch 23 - VAL - Batch 100 - Loss = 0.581 | Accuracy = 87.5%
544
+ 2023-01-04 20:14:17,714 - INFO - Epoch 23 - VAL - Batch 200 - Loss = 0.282 | Accuracy = 90.625%
545
+ 2023-01-04 20:14:22,450 - INFO - Epoch 23 - VAL - Batch 300 - Loss = 0.155 | Accuracy = 96.875%
546
+ 2023-01-04 20:14:26,383 - INFO - TRAIN phase
547
+ 2023-01-04 20:14:26,450 - INFO - Epoch 24 - TRAIN - Batch 0 - Loss = 0.338 | Accuracy = 90.625%
548
+ 2023-01-04 20:14:31,971 - INFO - Epoch 24 - TRAIN - Batch 100 - Loss = 0.622 | Accuracy = 84.375%
549
+ 2023-01-04 20:14:37,190 - INFO - Epoch 24 - TRAIN - Batch 200 - Loss = 0.82 | Accuracy = 81.25%
550
+ 2023-01-04 20:14:42,348 - INFO - Epoch 24 - TRAIN - Batch 300 - Loss = 0.094 | Accuracy = 96.875%
551
+ 2023-01-04 20:14:47,604 - INFO - Epoch 24 - TRAIN - Batch 400 - Loss = 0.371 | Accuracy = 87.5%
552
+ 2023-01-04 20:14:52,865 - INFO - Epoch 24 - TRAIN - Batch 500 - Loss = 0.284 | Accuracy = 87.5%
553
+ 2023-01-04 20:14:58,061 - INFO - Epoch 24 - TRAIN - Batch 600 - Loss = 0.276 | Accuracy = 84.375%
554
+ 2023-01-04 20:15:03,302 - INFO - Epoch 24 - TRAIN - Batch 700 - Loss = 0.51 | Accuracy = 81.25%
555
+ 2023-01-04 20:15:08,469 - INFO - Epoch 24 - TRAIN - Batch 800 - Loss = 0.226 | Accuracy = 87.5%
556
+ 2023-01-04 20:15:13,706 - INFO - Epoch 24 - TRAIN - Batch 900 - Loss = 0.351 | Accuracy = 87.5%
557
+ 2023-01-04 20:15:19,655 - INFO - Epoch 24 - TRAIN - Batch 1000 - Loss = 0.424 | Accuracy = 90.625%
558
+ 2023-01-04 20:15:25,450 - INFO - Epoch 24 - TRAIN - Batch 1100 - Loss = 0.178 | Accuracy = 93.75%
559
+ 2023-01-04 20:15:31,278 - INFO - Epoch 24 - TRAIN - Batch 1200 - Loss = 0.403 | Accuracy = 84.375%
560
+ 2023-01-04 20:15:37,135 - INFO - Epoch 24 - TRAIN - Batch 1300 - Loss = 0.347 | Accuracy = 87.5%
561
+ 2023-01-04 20:15:42,843 - INFO - Epoch 24 - TRAIN - Batch 1400 - Loss = 0.616 | Accuracy = 84.375%
562
+ 2023-01-04 20:15:48,689 - INFO - Epoch 24 - TRAIN - Batch 1500 - Loss = 0.474 | Accuracy = 81.25%
563
+ 2023-01-04 20:15:50,392 - INFO - VAL phase
564
+ 2023-01-04 20:15:50,454 - INFO - Epoch 24 - VAL - Batch 0 - Loss = 0.353 | Accuracy = 93.75%
565
+ 2023-01-04 20:15:55,259 - INFO - Epoch 24 - VAL - Batch 100 - Loss = 0.156 | Accuracy = 96.875%
566
+ 2023-01-04 20:16:00,121 - INFO - Epoch 24 - VAL - Batch 200 - Loss = 0.214 | Accuracy = 93.75%
567
+ 2023-01-04 20:16:05,278 - INFO - Epoch 24 - VAL - Batch 300 - Loss = 0.521 | Accuracy = 90.625%
568
+ 2023-01-05 02:12:51,799 - INFO - Initialized Digit model
569
+ 2023-01-05 02:12:51,987 - INFO -
570
+ Training details:
571
+ ------------------------
572
+ Model: HNet
573
+ Model Type: digit
574
+ Epochs: 25
575
+ Optimizer: SGD
576
+ Loss: CrossEntropyLoss
577
+ Learning Rate: 1e-05
578
+ Learning Rate Scheduler: <torch.optim.lr_scheduler.CyclicLR object at 0x000002412DDBFB20>
579
+ Batch Size: 32
580
+ Logging Interval: 100 batches
581
+ Train-dataset samples: 13600
582
+ Validation-dataset samples: 3400
583
+ -------------------------
584
+
585
+ 2023-01-05 02:12:51,987 - INFO - TRAIN phase
586
+ 2023-01-05 02:12:57,236 - INFO - Epoch 0 - TRAIN - Batch 0 - Loss = 2.263 | Accuracy = 6.25%
587
+ 2023-01-05 02:13:02,603 - INFO - Epoch 0 - TRAIN - Batch 100 - Loss = 2.242 | Accuracy = 18.75%
588
+ 2023-01-05 02:13:09,056 - INFO - Epoch 0 - TRAIN - Batch 200 - Loss = 2.28 | Accuracy = 9.375%
589
+ 2023-01-05 02:13:14,639 - INFO - Epoch 0 - TRAIN - Batch 300 - Loss = 2.155 | Accuracy = 31.25%
590
+ 2023-01-05 02:13:22,413 - INFO - Epoch 0 - TRAIN - Batch 400 - Loss = 2.125 | Accuracy = 40.625%
591
+ 2023-01-05 02:13:24,303 - INFO - VAL phase
592
+ 2023-01-05 02:13:24,365 - INFO - Epoch 0 - VAL - Batch 0 - Loss = 2.088 | Accuracy = 59.375%
593
+ 2023-01-05 02:13:30,288 - INFO - Epoch 0 - VAL - Batch 100 - Loss = 2.115 | Accuracy = 50.0%
594
+ 2023-01-05 02:13:30,618 - INFO - TRAIN phase
595
+ 2023-01-05 02:13:30,713 - INFO - Epoch 1 - TRAIN - Batch 0 - Loss = 2.139 | Accuracy = 31.25%
596
+ 2023-01-05 02:13:36,240 - INFO - Epoch 1 - TRAIN - Batch 100 - Loss = 1.543 | Accuracy = 65.625%
597
+ 2023-01-05 02:13:41,500 - INFO - Epoch 1 - TRAIN - Batch 200 - Loss = 1.224 | Accuracy = 65.625%
598
+ 2023-01-05 02:13:46,586 - INFO - Epoch 1 - TRAIN - Batch 300 - Loss = 0.925 | Accuracy = 75.0%
599
+ 2023-01-05 02:13:51,707 - INFO - Epoch 1 - TRAIN - Batch 400 - Loss = 0.788 | Accuracy = 75.0%
600
+ 2023-01-05 02:13:52,955 - INFO - VAL phase
601
+ 2023-01-05 02:13:53,017 - INFO - Epoch 1 - VAL - Batch 0 - Loss = 0.825 | Accuracy = 75.0%
602
+ 2023-01-05 02:13:58,185 - INFO - Epoch 1 - VAL - Batch 100 - Loss = 0.734 | Accuracy = 75.0%
603
+ 2023-01-05 02:13:58,491 - INFO - TRAIN phase
604
+ 2023-01-05 02:13:58,554 - INFO - Epoch 2 - TRAIN - Batch 0 - Loss = 0.785 | Accuracy = 81.25%
605
+ 2023-01-05 02:14:05,907 - INFO - Epoch 2 - TRAIN - Batch 100 - Loss = 0.795 | Accuracy = 71.875%
606
+ 2023-01-05 02:14:12,827 - INFO - Epoch 2 - TRAIN - Batch 200 - Loss = 1.233 | Accuracy = 62.5%
607
+ 2023-01-05 02:14:19,376 - INFO - Epoch 2 - TRAIN - Batch 300 - Loss = 0.467 | Accuracy = 87.5%
608
+ 2023-01-05 02:14:26,750 - INFO - Epoch 2 - TRAIN - Batch 400 - Loss = 0.497 | Accuracy = 90.625%
609
+ 2023-01-05 02:14:28,625 - INFO - VAL phase
610
+ 2023-01-05 02:14:28,683 - INFO - Epoch 2 - VAL - Batch 0 - Loss = 0.697 | Accuracy = 78.125%
611
+ 2023-01-05 02:14:34,362 - INFO - Epoch 2 - VAL - Batch 100 - Loss = 0.534 | Accuracy = 81.25%
612
+ 2023-01-05 02:14:34,648 - INFO - TRAIN phase
613
+ 2023-01-05 02:14:34,725 - INFO - Epoch 3 - TRAIN - Batch 0 - Loss = 0.621 | Accuracy = 81.25%
614
+ 2023-01-05 02:14:41,529 - INFO - Epoch 3 - TRAIN - Batch 100 - Loss = 0.864 | Accuracy = 71.875%
615
+ 2023-01-05 02:14:48,495 - INFO - Epoch 3 - TRAIN - Batch 200 - Loss = 0.587 | Accuracy = 81.25%
616
+ 2023-01-05 02:14:55,423 - INFO - Epoch 3 - TRAIN - Batch 300 - Loss = 0.789 | Accuracy = 68.75%
617
+ 2023-01-05 02:15:02,659 - INFO - Epoch 3 - TRAIN - Batch 400 - Loss = 0.347 | Accuracy = 90.625%
618
+ 2023-01-05 02:15:05,005 - INFO - VAL phase
619
+ 2023-01-05 02:15:05,066 - INFO - Epoch 3 - VAL - Batch 0 - Loss = 0.32 | Accuracy = 90.625%
620
+ 2023-01-05 02:15:11,962 - INFO - Epoch 3 - VAL - Batch 100 - Loss = 0.568 | Accuracy = 78.125%
621
+ 2023-01-05 02:15:12,287 - INFO - TRAIN phase
622
+ 2023-01-05 02:15:12,387 - INFO - Epoch 4 - TRAIN - Batch 0 - Loss = 0.389 | Accuracy = 87.5%
623
+ 2023-01-05 02:15:17,944 - INFO - Epoch 4 - TRAIN - Batch 100 - Loss = 0.737 | Accuracy = 75.0%
624
+ 2023-01-05 02:15:24,595 - INFO - Epoch 4 - TRAIN - Batch 200 - Loss = 0.476 | Accuracy = 84.375%
625
+ 2023-01-05 02:15:32,520 - INFO - Epoch 4 - TRAIN - Batch 300 - Loss = 0.597 | Accuracy = 78.125%
626
+ 2023-01-05 02:15:38,461 - INFO - Epoch 4 - TRAIN - Batch 400 - Loss = 0.471 | Accuracy = 84.375%
627
+ 2023-01-05 02:15:39,717 - INFO - VAL phase
628
+ 2023-01-05 02:15:39,777 - INFO - Epoch 4 - VAL - Batch 0 - Loss = 0.269 | Accuracy = 93.75%
629
+ 2023-01-05 02:15:44,843 - INFO - Epoch 4 - VAL - Batch 100 - Loss = 0.389 | Accuracy = 93.75%
630
+ 2023-01-05 02:15:45,123 - INFO - TRAIN phase
631
+ 2023-01-05 02:15:45,187 - INFO - Epoch 5 - TRAIN - Batch 0 - Loss = 0.735 | Accuracy = 84.375%
632
+ 2023-01-05 02:15:50,450 - INFO - Epoch 5 - TRAIN - Batch 100 - Loss = 0.396 | Accuracy = 84.375%
633
+ 2023-01-05 02:15:55,769 - INFO - Epoch 5 - TRAIN - Batch 200 - Loss = 0.449 | Accuracy = 84.375%
634
+ 2023-01-05 02:16:01,237 - INFO - Epoch 5 - TRAIN - Batch 300 - Loss = 0.353 | Accuracy = 87.5%
635
+ 2023-01-05 02:16:07,832 - INFO - Epoch 5 - TRAIN - Batch 400 - Loss = 0.564 | Accuracy = 78.125%
636
+ 2023-01-05 02:16:09,531 - INFO - VAL phase
637
+ 2023-01-05 02:16:09,587 - INFO - Epoch 5 - VAL - Batch 0 - Loss = 0.337 | Accuracy = 90.625%
638
+ 2023-01-05 02:16:15,092 - INFO - Epoch 5 - VAL - Batch 100 - Loss = 0.222 | Accuracy = 96.875%
639
+ 2023-01-05 02:16:15,387 - INFO - TRAIN phase
640
+ 2023-01-05 02:16:15,460 - INFO - Epoch 6 - TRAIN - Batch 0 - Loss = 0.47 | Accuracy = 87.5%
641
+ 2023-01-05 02:16:22,367 - INFO - Epoch 6 - TRAIN - Batch 100 - Loss = 0.589 | Accuracy = 90.625%
642
+ 2023-01-05 02:16:29,371 - INFO - Epoch 6 - TRAIN - Batch 200 - Loss = 0.419 | Accuracy = 87.5%
643
+ 2023-01-05 02:16:36,148 - INFO - Epoch 6 - TRAIN - Batch 300 - Loss = 0.197 | Accuracy = 93.75%
644
+ 2023-01-05 02:16:42,859 - INFO - Epoch 6 - TRAIN - Batch 400 - Loss = 0.201 | Accuracy = 90.625%
645
+ 2023-01-05 02:16:44,484 - INFO - VAL phase
646
+ 2023-01-05 02:16:44,543 - INFO - Epoch 6 - VAL - Batch 0 - Loss = 0.354 | Accuracy = 90.625%
647
+ 2023-01-05 02:16:49,849 - INFO - Epoch 6 - VAL - Batch 100 - Loss = 0.138 | Accuracy = 100.0%
648
+ 2023-01-05 02:16:50,149 - INFO - TRAIN phase
649
+ 2023-01-05 02:16:50,229 - INFO - Epoch 7 - TRAIN - Batch 0 - Loss = 0.309 | Accuracy = 90.625%
650
+ 2023-01-05 02:16:57,530 - INFO - Epoch 7 - TRAIN - Batch 100 - Loss = 0.153 | Accuracy = 96.875%
651
+ 2023-01-05 02:17:04,317 - INFO - Epoch 7 - TRAIN - Batch 200 - Loss = 0.128 | Accuracy = 96.875%
652
+ 2023-01-05 02:17:11,037 - INFO - Epoch 7 - TRAIN - Batch 300 - Loss = 0.463 | Accuracy = 84.375%
653
+ 2023-01-05 02:17:17,910 - INFO - Epoch 7 - TRAIN - Batch 400 - Loss = 0.44 | Accuracy = 84.375%
654
+ 2023-01-05 02:17:19,521 - INFO - VAL phase
655
+ 2023-01-05 02:17:19,588 - INFO - Epoch 7 - VAL - Batch 0 - Loss = 0.119 | Accuracy = 100.0%
656
+ 2023-01-05 02:17:25,217 - INFO - Epoch 7 - VAL - Batch 100 - Loss = 0.162 | Accuracy = 96.875%
657
+ 2023-01-05 02:17:25,517 - INFO - TRAIN phase
658
+ 2023-01-05 02:17:25,593 - INFO - Epoch 8 - TRAIN - Batch 0 - Loss = 0.119 | Accuracy = 96.875%
659
+ 2023-01-05 02:17:32,328 - INFO - Epoch 8 - TRAIN - Batch 100 - Loss = 0.362 | Accuracy = 87.5%
660
+ 2023-01-05 02:17:39,141 - INFO - Epoch 8 - TRAIN - Batch 200 - Loss = 0.204 | Accuracy = 87.5%
661
+ 2023-01-05 02:17:45,858 - INFO - Epoch 8 - TRAIN - Batch 300 - Loss = 0.259 | Accuracy = 90.625%
662
+ 2023-01-05 02:17:52,601 - INFO - Epoch 8 - TRAIN - Batch 400 - Loss = 0.245 | Accuracy = 93.75%
663
+ 2023-01-05 02:17:54,217 - INFO - VAL phase
664
+ 2023-01-05 02:17:54,281 - INFO - Epoch 8 - VAL - Batch 0 - Loss = 0.171 | Accuracy = 93.75%
665
+ 2023-01-05 02:17:59,745 - INFO - Epoch 8 - VAL - Batch 100 - Loss = 0.065 | Accuracy = 100.0%
666
+ 2023-01-05 02:18:00,050 - INFO - TRAIN phase
667
+ 2023-01-05 02:18:00,138 - INFO - Epoch 9 - TRAIN - Batch 0 - Loss = 0.083 | Accuracy = 96.875%
668
+ 2023-01-05 02:18:07,297 - INFO - Epoch 9 - TRAIN - Batch 100 - Loss = 0.555 | Accuracy = 90.625%
669
+ 2023-01-05 02:18:13,999 - INFO - Epoch 9 - TRAIN - Batch 200 - Loss = 0.11 | Accuracy = 96.875%
670
+ 2023-01-05 02:18:20,827 - INFO - Epoch 9 - TRAIN - Batch 300 - Loss = 0.047 | Accuracy = 100.0%
671
+ 2023-01-05 02:18:27,618 - INFO - Epoch 9 - TRAIN - Batch 400 - Loss = 0.236 | Accuracy = 96.875%
672
+ 2023-01-05 02:18:29,228 - INFO - VAL phase
673
+ 2023-01-05 02:18:29,296 - INFO - Epoch 9 - VAL - Batch 0 - Loss = 0.041 | Accuracy = 100.0%
674
+ 2023-01-05 02:18:34,680 - INFO - Epoch 9 - VAL - Batch 100 - Loss = 0.08 | Accuracy = 100.0%
675
+ 2023-01-05 02:18:34,950 - INFO - TRAIN phase
676
+ 2023-01-05 02:18:35,029 - INFO - Epoch 10 - TRAIN - Batch 0 - Loss = 0.183 | Accuracy = 96.875%
677
+ 2023-01-05 02:18:41,673 - INFO - Epoch 10 - TRAIN - Batch 100 - Loss = 0.31 | Accuracy = 90.625%
678
+ 2023-01-05 02:18:48,329 - INFO - Epoch 10 - TRAIN - Batch 200 - Loss = 0.086 | Accuracy = 100.0%
679
+ 2023-01-05 02:18:55,169 - INFO - Epoch 10 - TRAIN - Batch 300 - Loss = 0.172 | Accuracy = 90.625%
680
+ 2023-01-05 02:19:01,831 - INFO - Epoch 10 - TRAIN - Batch 400 - Loss = 0.083 | Accuracy = 96.875%
681
+ 2023-01-05 02:19:03,448 - INFO - VAL phase
682
+ 2023-01-05 02:19:03,505 - INFO - Epoch 10 - VAL - Batch 0 - Loss = 0.103 | Accuracy = 96.875%
683
+ 2023-01-05 02:19:09,054 - INFO - Epoch 10 - VAL - Batch 100 - Loss = 0.121 | Accuracy = 96.875%
684
+ 2023-01-05 02:19:09,410 - INFO - TRAIN phase
685
+ 2023-01-05 02:19:09,531 - INFO - Epoch 11 - TRAIN - Batch 0 - Loss = 0.155 | Accuracy = 93.75%
686
+ 2023-01-05 02:19:17,030 - INFO - Epoch 11 - TRAIN - Batch 100 - Loss = 0.185 | Accuracy = 93.75%
687
+ 2023-01-05 02:19:24,134 - INFO - Epoch 11 - TRAIN - Batch 200 - Loss = 0.09 | Accuracy = 96.875%
688
+ 2023-01-05 02:19:30,872 - INFO - Epoch 11 - TRAIN - Batch 300 - Loss = 0.117 | Accuracy = 96.875%
689
+ 2023-01-05 02:19:37,546 - INFO - Epoch 11 - TRAIN - Batch 400 - Loss = 0.06 | Accuracy = 100.0%
690
+ 2023-01-05 02:19:39,131 - INFO - VAL phase
691
+ 2023-01-05 02:19:39,188 - INFO - Epoch 11 - VAL - Batch 0 - Loss = 0.066 | Accuracy = 96.875%
692
+ 2023-01-05 02:19:44,612 - INFO - Epoch 11 - VAL - Batch 100 - Loss = 0.225 | Accuracy = 93.75%
693
+ 2023-01-05 02:19:44,915 - INFO - TRAIN phase
694
+ 2023-01-05 02:19:44,987 - INFO - Epoch 12 - TRAIN - Batch 0 - Loss = 0.153 | Accuracy = 93.75%
695
+ 2023-01-05 02:19:51,778 - INFO - Epoch 12 - TRAIN - Batch 100 - Loss = 0.217 | Accuracy = 93.75%
696
+ 2023-01-05 02:19:58,458 - INFO - Epoch 12 - TRAIN - Batch 200 - Loss = 0.023 | Accuracy = 100.0%
697
+ 2023-01-05 02:20:05,163 - INFO - Epoch 12 - TRAIN - Batch 300 - Loss = 0.239 | Accuracy = 90.625%
698
+ 2023-01-05 02:20:11,837 - INFO - Epoch 12 - TRAIN - Batch 400 - Loss = 0.208 | Accuracy = 90.625%
699
+ 2023-01-05 02:20:13,471 - INFO - VAL phase
700
+ 2023-01-05 02:20:13,539 - INFO - Epoch 12 - VAL - Batch 0 - Loss = 0.119 | Accuracy = 96.875%
701
+ 2023-01-05 02:20:19,061 - INFO - Epoch 12 - VAL - Batch 100 - Loss = 0.096 | Accuracy = 93.75%
702
+ 2023-01-05 02:20:19,359 - INFO - TRAIN phase
703
+ 2023-01-05 02:20:19,447 - INFO - Epoch 13 - TRAIN - Batch 0 - Loss = 0.249 | Accuracy = 90.625%
704
+ 2023-01-05 02:20:24,932 - INFO - Epoch 13 - TRAIN - Batch 100 - Loss = 0.312 | Accuracy = 96.875%
705
+ 2023-01-05 02:20:30,033 - INFO - Epoch 13 - TRAIN - Batch 200 - Loss = 0.239 | Accuracy = 90.625%
706
+ 2023-01-05 02:20:35,143 - INFO - Epoch 13 - TRAIN - Batch 300 - Loss = 0.051 | Accuracy = 100.0%
707
+ 2023-01-05 02:20:40,288 - INFO - Epoch 13 - TRAIN - Batch 400 - Loss = 0.063 | Accuracy = 100.0%
708
+ 2023-01-05 02:20:41,525 - INFO - VAL phase
709
+ 2023-01-05 02:20:41,580 - INFO - Epoch 13 - VAL - Batch 0 - Loss = 0.042 | Accuracy = 100.0%
710
+ 2023-01-05 02:20:46,552 - INFO - Epoch 13 - VAL - Batch 100 - Loss = 0.189 | Accuracy = 93.75%
711
+ 2023-01-05 02:20:46,803 - INFO - TRAIN phase
712
+ 2023-01-05 02:20:46,868 - INFO - Epoch 14 - TRAIN - Batch 0 - Loss = 0.131 | Accuracy = 96.875%
713
+ 2023-01-05 02:20:52,046 - INFO - Epoch 14 - TRAIN - Batch 100 - Loss = 0.122 | Accuracy = 96.875%
714
+ 2023-01-05 02:20:57,355 - INFO - Epoch 14 - TRAIN - Batch 200 - Loss = 0.028 | Accuracy = 100.0%
715
+ 2023-01-05 02:21:02,607 - INFO - Epoch 14 - TRAIN - Batch 300 - Loss = 0.38 | Accuracy = 84.375%
716
+ 2023-01-05 02:21:07,999 - INFO - Epoch 14 - TRAIN - Batch 400 - Loss = 0.125 | Accuracy = 96.875%
717
+ 2023-01-05 02:21:09,409 - INFO - VAL phase
718
+ 2023-01-05 02:21:09,465 - INFO - Epoch 14 - VAL - Batch 0 - Loss = 0.155 | Accuracy = 96.875%
719
+ 2023-01-05 02:21:14,908 - INFO - Epoch 14 - VAL - Batch 100 - Loss = 0.094 | Accuracy = 93.75%
720
+ 2023-01-05 02:21:15,216 - INFO - TRAIN phase
721
+ 2023-01-05 02:21:15,316 - INFO - Epoch 15 - TRAIN - Batch 0 - Loss = 0.094 | Accuracy = 96.875%
722
+ 2023-01-05 02:21:22,818 - INFO - Epoch 15 - TRAIN - Batch 100 - Loss = 0.025 | Accuracy = 100.0%
723
+ 2023-01-05 02:21:29,898 - INFO - Epoch 15 - TRAIN - Batch 200 - Loss = 0.184 | Accuracy = 93.75%
724
+ 2023-01-05 02:21:36,685 - INFO - Epoch 15 - TRAIN - Batch 300 - Loss = 0.078 | Accuracy = 96.875%
725
+ 2023-01-05 02:21:43,467 - INFO - Epoch 15 - TRAIN - Batch 400 - Loss = 0.124 | Accuracy = 96.875%
726
+ 2023-01-05 02:21:45,083 - INFO - VAL phase
727
+ 2023-01-05 02:21:45,140 - INFO - Epoch 15 - VAL - Batch 0 - Loss = 0.068 | Accuracy = 96.875%
728
+ 2023-01-05 02:21:50,597 - INFO - Epoch 15 - VAL - Batch 100 - Loss = 0.027 | Accuracy = 100.0%
729
+ 2023-01-05 02:21:50,896 - INFO - TRAIN phase
730
+ 2023-01-05 02:21:50,977 - INFO - Epoch 16 - TRAIN - Batch 0 - Loss = 0.261 | Accuracy = 93.75%
731
+ 2023-01-05 02:21:57,785 - INFO - Epoch 16 - TRAIN - Batch 100 - Loss = 0.116 | Accuracy = 93.75%
732
+ 2023-01-05 02:22:04,469 - INFO - Epoch 16 - TRAIN - Batch 200 - Loss = 0.089 | Accuracy = 93.75%
733
+ 2023-01-05 02:22:11,198 - INFO - Epoch 16 - TRAIN - Batch 300 - Loss = 0.184 | Accuracy = 96.875%
734
+ 2023-01-05 02:22:17,933 - INFO - Epoch 16 - TRAIN - Batch 400 - Loss = 0.05 | Accuracy = 100.0%
735
+ 2023-01-05 02:22:19,556 - INFO - VAL phase
736
+ 2023-01-05 02:22:19,615 - INFO - Epoch 16 - VAL - Batch 0 - Loss = 0.116 | Accuracy = 96.875%
737
+ 2023-01-05 02:22:25,161 - INFO - Epoch 16 - VAL - Batch 100 - Loss = 0.038 | Accuracy = 100.0%
738
+ 2023-01-05 02:22:25,461 - INFO - TRAIN phase
739
+ 2023-01-05 02:22:25,552 - INFO - Epoch 17 - TRAIN - Batch 0 - Loss = 0.031 | Accuracy = 100.0%
740
+ 2023-01-05 02:22:32,825 - INFO - Epoch 17 - TRAIN - Batch 100 - Loss = 0.15 | Accuracy = 93.75%
741
+ 2023-01-05 02:22:39,548 - INFO - Epoch 17 - TRAIN - Batch 200 - Loss = 0.433 | Accuracy = 90.625%
742
+ 2023-01-05 02:22:46,312 - INFO - Epoch 17 - TRAIN - Batch 300 - Loss = 0.037 | Accuracy = 100.0%
743
+ 2023-01-05 02:22:53,015 - INFO - Epoch 17 - TRAIN - Batch 400 - Loss = 0.192 | Accuracy = 93.75%
744
+ 2023-01-05 02:22:54,653 - INFO - VAL phase
745
+ 2023-01-05 02:22:54,699 - INFO - Epoch 17 - VAL - Batch 0 - Loss = 0.043 | Accuracy = 100.0%
746
+ 2023-01-05 02:23:00,109 - INFO - Epoch 17 - VAL - Batch 100 - Loss = 0.037 | Accuracy = 100.0%
747
+ 2023-01-05 02:23:00,385 - INFO - TRAIN phase
748
+ 2023-01-05 02:23:00,458 - INFO - Epoch 18 - TRAIN - Batch 0 - Loss = 0.253 | Accuracy = 93.75%
749
+ 2023-01-05 02:23:07,206 - INFO - Epoch 18 - TRAIN - Batch 100 - Loss = 0.325 | Accuracy = 93.75%
750
+ 2023-01-05 02:23:13,836 - INFO - Epoch 18 - TRAIN - Batch 200 - Loss = 0.015 | Accuracy = 100.0%
751
+ 2023-01-05 02:23:20,695 - INFO - Epoch 18 - TRAIN - Batch 300 - Loss = 0.14 | Accuracy = 90.625%
752
+ 2023-01-05 02:23:27,388 - INFO - Epoch 18 - TRAIN - Batch 400 - Loss = 0.086 | Accuracy = 100.0%
753
+ 2023-01-05 02:23:29,040 - INFO - VAL phase
754
+ 2023-01-05 02:23:29,101 - INFO - Epoch 18 - VAL - Batch 0 - Loss = 0.132 | Accuracy = 93.75%
755
+ 2023-01-05 02:23:34,433 - INFO - Epoch 18 - VAL - Batch 100 - Loss = 0.163 | Accuracy = 93.75%
756
+ 2023-01-05 02:23:34,719 - INFO - TRAIN phase
757
+ 2023-01-05 02:23:34,792 - INFO - Epoch 19 - TRAIN - Batch 0 - Loss = 0.053 | Accuracy = 96.875%
758
+ 2023-01-05 02:23:41,450 - INFO - Epoch 19 - TRAIN - Batch 100 - Loss = 0.253 | Accuracy = 96.875%
759
+ 2023-01-05 02:23:48,095 - INFO - Epoch 19 - TRAIN - Batch 200 - Loss = 0.107 | Accuracy = 96.875%
760
+ 2023-01-05 02:23:54,842 - INFO - Epoch 19 - TRAIN - Batch 300 - Loss = 0.004 | Accuracy = 100.0%
761
+ 2023-01-05 02:24:01,659 - INFO - Epoch 19 - TRAIN - Batch 400 - Loss = 0.068 | Accuracy = 96.875%
762
+ 2023-01-05 02:24:03,504 - INFO - VAL phase
763
+ 2023-01-05 02:24:03,567 - INFO - Epoch 19 - VAL - Batch 0 - Loss = 0.677 | Accuracy = 93.75%
764
+ 2023-01-05 02:24:09,055 - INFO - Epoch 19 - VAL - Batch 100 - Loss = 0.015 | Accuracy = 100.0%
765
+ 2023-01-05 02:24:09,371 - INFO - TRAIN phase
766
+ 2023-01-05 02:24:09,463 - INFO - Epoch 20 - TRAIN - Batch 0 - Loss = 0.067 | Accuracy = 100.0%
767
+ 2023-01-05 02:24:16,520 - INFO - Epoch 20 - TRAIN - Batch 100 - Loss = 0.07 | Accuracy = 100.0%
768
+ 2023-01-05 02:24:23,168 - INFO - Epoch 20 - TRAIN - Batch 200 - Loss = 0.081 | Accuracy = 100.0%
769
+ 2023-01-05 02:24:29,900 - INFO - Epoch 20 - TRAIN - Batch 300 - Loss = 0.129 | Accuracy = 96.875%
770
+ 2023-01-05 02:24:37,273 - INFO - Epoch 20 - TRAIN - Batch 400 - Loss = 0.056 | Accuracy = 96.875%
771
+ 2023-01-05 02:24:39,092 - INFO - VAL phase
772
+ 2023-01-05 02:24:39,166 - INFO - Epoch 20 - VAL - Batch 0 - Loss = 0.357 | Accuracy = 93.75%
773
+ 2023-01-05 02:24:46,786 - INFO - Epoch 20 - VAL - Batch 100 - Loss = 0.07 | Accuracy = 100.0%
774
+ 2023-01-05 02:24:47,074 - INFO - TRAIN phase
775
+ 2023-01-05 02:24:47,144 - INFO - Epoch 21 - TRAIN - Batch 0 - Loss = 0.116 | Accuracy = 96.875%
776
+ 2023-01-05 02:24:54,208 - INFO - Epoch 21 - TRAIN - Batch 100 - Loss = 0.117 | Accuracy = 93.75%
777
+ 2023-01-05 02:25:01,012 - INFO - Epoch 21 - TRAIN - Batch 200 - Loss = 0.111 | Accuracy = 96.875%
778
+ 2023-01-05 02:25:07,726 - INFO - Epoch 21 - TRAIN - Batch 300 - Loss = 0.03 | Accuracy = 100.0%
779
+ 2023-01-05 02:25:14,537 - INFO - Epoch 21 - TRAIN - Batch 400 - Loss = 0.277 | Accuracy = 90.625%
780
+ 2023-01-05 02:25:16,150 - INFO - VAL phase
781
+ 2023-01-05 02:25:16,204 - INFO - Epoch 21 - VAL - Batch 0 - Loss = 0.04 | Accuracy = 100.0%
782
+ 2023-01-05 02:25:21,644 - INFO - Epoch 21 - VAL - Batch 100 - Loss = 0.029 | Accuracy = 100.0%
783
+ 2023-01-05 02:25:21,947 - INFO - TRAIN phase
784
+ 2023-01-05 02:25:22,022 - INFO - Epoch 22 - TRAIN - Batch 0 - Loss = 0.024 | Accuracy = 96.875%
785
+ 2023-01-05 02:25:28,757 - INFO - Epoch 22 - TRAIN - Batch 100 - Loss = 0.025 | Accuracy = 100.0%
786
+ 2023-01-05 02:25:35,481 - INFO - Epoch 22 - TRAIN - Batch 200 - Loss = 0.16 | Accuracy = 93.75%
787
+ 2023-01-05 02:25:42,234 - INFO - Epoch 22 - TRAIN - Batch 300 - Loss = 0.473 | Accuracy = 87.5%
788
+ 2023-01-05 02:25:48,980 - INFO - Epoch 22 - TRAIN - Batch 400 - Loss = 0.032 | Accuracy = 100.0%
789
+ 2023-01-05 02:25:50,590 - INFO - VAL phase
790
+ 2023-01-05 02:25:50,654 - INFO - Epoch 22 - VAL - Batch 0 - Loss = 0.045 | Accuracy = 96.875%
791
+ 2023-01-05 02:25:56,118 - INFO - Epoch 22 - VAL - Batch 100 - Loss = 0.115 | Accuracy = 93.75%
792
+ 2023-01-05 02:25:56,391 - INFO - TRAIN phase
793
+ 2023-01-05 02:25:56,491 - INFO - Epoch 23 - TRAIN - Batch 0 - Loss = 0.066 | Accuracy = 96.875%
794
+ 2023-01-05 02:26:03,868 - INFO - Epoch 23 - TRAIN - Batch 100 - Loss = 0.141 | Accuracy = 93.75%
795
+ 2023-01-05 02:26:10,572 - INFO - Epoch 23 - TRAIN - Batch 200 - Loss = 0.083 | Accuracy = 96.875%
796
+ 2023-01-05 02:26:17,329 - INFO - Epoch 23 - TRAIN - Batch 300 - Loss = 0.114 | Accuracy = 96.875%
797
+ 2023-01-05 02:26:24,090 - INFO - Epoch 23 - TRAIN - Batch 400 - Loss = 0.304 | Accuracy = 96.875%
798
+ 2023-01-05 02:26:25,694 - INFO - VAL phase
799
+ 2023-01-05 02:26:25,754 - INFO - Epoch 23 - VAL - Batch 0 - Loss = 0.045 | Accuracy = 96.875%
800
+ 2023-01-05 02:26:31,192 - INFO - Epoch 23 - VAL - Batch 100 - Loss = 0.14 | Accuracy = 96.875%
801
+ 2023-01-05 02:26:31,466 - INFO - TRAIN phase
802
+ 2023-01-05 02:26:31,540 - INFO - Epoch 24 - TRAIN - Batch 0 - Loss = 0.019 | Accuracy = 100.0%
803
+ 2023-01-05 02:26:38,284 - INFO - Epoch 24 - TRAIN - Batch 100 - Loss = 0.065 | Accuracy = 96.875%
804
+ 2023-01-05 02:26:45,000 - INFO - Epoch 24 - TRAIN - Batch 200 - Loss = 0.039 | Accuracy = 100.0%
805
+ 2023-01-05 02:26:51,860 - INFO - Epoch 24 - TRAIN - Batch 300 - Loss = 0.154 | Accuracy = 93.75%
806
+ 2023-01-05 02:26:58,681 - INFO - Epoch 24 - TRAIN - Batch 400 - Loss = 0.092 | Accuracy = 96.875%
807
+ 2023-01-05 02:27:00,306 - INFO - VAL phase
808
+ 2023-01-05 02:27:00,361 - INFO - Epoch 24 - VAL - Batch 0 - Loss = 0.429 | Accuracy = 90.625%
809
+ 2023-01-05 02:27:05,841 - INFO - Epoch 24 - VAL - Batch 100 - Loss = 0.056 | Accuracy = 100.0%
src/train.py ADDED
@@ -0,0 +1,265 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import torch
2
+ import torch.nn as nn
3
+ from torch.optim import SGD, lr_scheduler
4
+ from torch.nn import CrossEntropyLoss
5
+ from torch.utils.data import DataLoader, random_split
6
+ from torchvision.datasets import ImageFolder
7
+ from model import HNet, ResNet18
8
+ import config as CFG
9
+ from tqdm.auto import tqdm
10
+ from prettytable import PrettyTable
11
+ from argparse import ArgumentParser
12
+ from copy import deepcopy
13
+ from typing import Dict
14
+ import time
15
+ import logging
16
+ import sys
17
+ from data import transforms
18
+
19
+ # check is models folder exists
20
+ (CFG.BASE_PATH / "models").mkdir(exist_ok=True)
21
+
22
+
23
+ # Set up logger
24
+ logging.basicConfig(
25
+ filename="train.log",
26
+ format="%(asctime)s - %(levelname)s - %(message)s",
27
+ level=logging.INFO,
28
+ filemode="a",
29
+ )
30
+
31
+
32
+ best_acc = 0.0
33
+
34
+
35
+ def run_one_epoch(
36
+ epoch: int,
37
+ ds_sizes: Dict[str, int],
38
+ dataloaders: Dict[str, DataLoader],
39
+ model: nn.Module,
40
+ optimizer: torch.optim.Optimizer,
41
+ loss: nn.Module,
42
+ scheduler: torch.optim.lr_scheduler,
43
+ ):
44
+ """
45
+ Run one complete train-val loop
46
+
47
+ Parameter
48
+ ---------
49
+
50
+ ds_sizes: Dictionary containing dataset sizes
51
+ dataloaders: Dictionary containing dataloaders
52
+ model: The model
53
+ optimizer: The optimizer
54
+ loss: The loss
55
+
56
+ Returns
57
+ -------
58
+
59
+ metrics: Dictionary containing Train(loss/accuracy) &
60
+ Validation(loss/accuracy)
61
+
62
+ """
63
+ global best_acc
64
+
65
+ metrics = {}
66
+
67
+ for phase in ["train", "val"]:
68
+ logging.info(f"{phase.upper()} phase")
69
+
70
+ if phase == "train":
71
+ model.train()
72
+ else:
73
+ model.eval()
74
+
75
+ avg_loss = 0
76
+ running_corrects = 0
77
+
78
+ for batch_idx, (images, labels) in enumerate(
79
+ tqdm(dataloaders[phase], total=len(dataloaders[phase]))
80
+ ):
81
+
82
+ images = images.to(CFG.DEVICE)
83
+ labels = labels.to(CFG.DEVICE)
84
+
85
+ # Zero the gradients
86
+ optimizer.zero_grad()
87
+
88
+ # Track history if in phase == "train"
89
+ with torch.set_grad_enabled(phase == "train"):
90
+ outputs = model(images)
91
+ _, preds = torch.max(outputs, 1)
92
+ loss = criterion(outputs, labels)
93
+
94
+ if phase == "train":
95
+ loss.backward()
96
+ optimizer.step()
97
+
98
+ avg_loss += loss.item() * images.size(0)
99
+ running_corrects += torch.sum(preds == labels)
100
+
101
+ if batch_idx % CFG.INTERVAL == 0:
102
+ corrects = torch.sum(preds == labels)
103
+
104
+ logging.info(
105
+ f"Epoch {epoch} - {phase.upper()} - Batch {batch_idx} - Loss = {round(loss.item(), 3)} | Accuracy = {100 * corrects/CFG.BATCH_SIZE}%"
106
+ )
107
+
108
+ epoch_loss = avg_loss / ds_sizes[phase]
109
+ epoch_acc = running_corrects.double() / ds_sizes[phase]
110
+
111
+ # step the scheduler
112
+ if phase == "train":
113
+ scheduler.step()
114
+
115
+ # save best model wts
116
+ if phase == "val" and epoch_acc > best_acc:
117
+ best_acc = epoch_acc
118
+ best_model_wts = deepcopy(model.state_dict())
119
+ torch.save(best_model_wts, CFG.BEST_MODEL_PATH)
120
+
121
+ # Metrics tracking
122
+ if phase == "train":
123
+ metrics["train_loss"] = round(epoch_loss, 3)
124
+ metrics["train_acc"] = round(100 * epoch_acc.item(), 3)
125
+ else:
126
+ metrics["val_loss"] = round(epoch_loss, 3)
127
+ metrics["val_acc"] = round(100 * epoch_acc.item(), 3)
128
+
129
+ return metrics
130
+
131
+
132
+ def train(dataloaders, ds_sizes, model, optimizer, criterion, scheduler):
133
+ for epoch in range(CFG.EPOCHS):
134
+
135
+ start = time.time()
136
+
137
+ metrics = run_one_epoch(
138
+ epoch=epoch,
139
+ ds_sizes=ds_sizes,
140
+ dataloaders=dataloaders,
141
+ model=model,
142
+ optimizer=optimizer,
143
+ loss=criterion,
144
+ scheduler=scheduler,
145
+ )
146
+
147
+ end = time.time() - start
148
+
149
+ print(f"Epoch completed in: {round(end/60, 3)} mins")
150
+
151
+ table.add_row(
152
+ row=[
153
+ epoch + 1,
154
+ metrics["train_loss"],
155
+ metrics["train_acc"],
156
+ metrics["val_loss"],
157
+ metrics["val_acc"],
158
+ ]
159
+ )
160
+ print(table)
161
+
162
+ # Write results to file
163
+ with open("results.txt", "w") as f:
164
+ results = table.get_string()
165
+ f.write(results)
166
+
167
+
168
+ if __name__ == "__main__":
169
+
170
+ TRAIN_PATH, TEST_PATH, BEST_MODEL = "", "", ""
171
+
172
+ parser = ArgumentParser(description="Train model for Hindi Character Recognition")
173
+ parser.add_argument(
174
+ "--epochs", type=int, help="number of epochs", default=CFG.EPOCHS
175
+ )
176
+ parser.add_argument("--lr", type=float, help="learning rate", default=CFG.LR)
177
+ parser.add_argument(
178
+ "--model_type",
179
+ type=str,
180
+ help="Type of model (vyanjan/digit)",
181
+ default="vyanjan",
182
+ )
183
+
184
+ args = parser.parse_args()
185
+
186
+ if args.model_type == "digit":
187
+ model = HNet(num_classes=10)
188
+ logging.info("Initialized Digit model")
189
+ TRAIN_PATH = CFG.TRAIN_DIGIT_PATH
190
+ CFG.BEST_MODEL_PATH = CFG.BEST_MODEL_DIGIT
191
+ else:
192
+ model = HNet(num_classes=36)
193
+ logging.info("Initialized Vyanjan model")
194
+ TRAIN_PATH = CFG.TRAIN_VYANJAN_PATH
195
+ CFG.BEST_MODEL_PATH = CFG.BEST_MODEL_VYANJAN
196
+
197
+ # creating the datasets
198
+ train_ds = ImageFolder(root=TRAIN_PATH, transform=transforms["train"])
199
+
200
+ # Train/val splitting
201
+ lengths = [int(len(train_ds) * 0.8), len(train_ds) - int(len(train_ds) * 0.8)]
202
+ train_ds, val_ds = random_split(dataset=train_ds, lengths=lengths)
203
+
204
+ # creating the dataloaders
205
+ train_dl = DataLoader(dataset=train_ds, batch_size=CFG.BATCH_SIZE, shuffle=True)
206
+ val_dl = DataLoader(dataset=val_ds, batch_size=CFG.BATCH_SIZE)
207
+
208
+ if len(sys.argv) > 1:
209
+ CFG.EPOCHS = args.epochs
210
+ CFG.LR = args.lr
211
+
212
+ # table
213
+ table = PrettyTable(
214
+ field_names=["Epoch", "Train Loss", "Train Acc", "Val Loss", "Val Acc"]
215
+ )
216
+
217
+ # the model
218
+ model.to(CFG.DEVICE)
219
+
220
+ # Setting up optimizer and loss
221
+ optimizer = SGD(model.parameters(), lr=CFG.LR)
222
+ criterion = CrossEntropyLoss()
223
+
224
+ scheduler = lr_scheduler.CyclicLR(
225
+ optimizer=optimizer, base_lr=1e-5, max_lr=0.1, verbose=True
226
+ )
227
+
228
+ dataloaders = {"train": train_dl, "val": val_dl}
229
+ ds_sizes = {"train": len(train_ds), "val": len(val_ds)}
230
+
231
+ detail = f"""
232
+ Training details:
233
+ ------------------------
234
+ Model: {model._get_name()}
235
+ Model Type: {args.model_type}
236
+ Epochs: {CFG.EPOCHS}
237
+ Optimizer: {type(optimizer).__name__}
238
+ Loss: {criterion._get_name()}
239
+ Learning Rate: {CFG.LR}
240
+ Learning Rate Scheduler: {scheduler.__str__()}
241
+ Batch Size: {CFG.BATCH_SIZE}
242
+ Logging Interval: {CFG.INTERVAL} batches
243
+ Train-dataset samples: {len(train_ds)}
244
+ Validation-dataset samples: {len(val_ds)}
245
+ -------------------------
246
+ """
247
+
248
+ print(detail)
249
+
250
+ logging.info(detail)
251
+
252
+ start_train = time.time()
253
+
254
+ train(
255
+ dataloaders=dataloaders,
256
+ ds_sizes=ds_sizes,
257
+ model=model,
258
+ optimizer=optimizer,
259
+ criterion=criterion,
260
+ scheduler=scheduler,
261
+ )
262
+
263
+ end_train = time.time() - start_train
264
+
265
+ print(f"Training completed in: {round(end_train/60, 3)} mins")
src/vyanjan_mapping.png ADDED