-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathplot_loss_metric.py
91 lines (76 loc) · 4.28 KB
/
plot_loss_metric.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
import re
import matplotlib.pyplot as plt
# Define extraction patterns
pattern_train = re.compile(r'Epoch \[(\d+)/\d+\] \| Step \[(\d+)/\d+\] \| Train Loss: ([-\d.]+)')
pattern_valid = re.compile(r'Epoch \[(\d+)/\d+\] \| Step \[(\d+)/\d+\] \| Valid Loss: ([-\d.]+)')
pattern_metric = re.compile(r'Epoch \[(\d+)/\d+\] \| Step \[(\d+)/\d+\] \| Valid Loss: [-\d.]+, Metric: ([\d.]+)')
# Function to extract data from a file
def extract_data_from_file(filepath):
epochs = []
train_losses = []
valid_epochs_loss = []
valid_losses = []
valid_epochs = []
valid_metrics = []
with open(filepath, "r") as file:
lines = file.readlines()
for line in lines:
match_train = pattern_train.search(line)
match_valid = pattern_valid.search(line)
match_metric = pattern_metric.search(line)
if match_train:
epoch, _, loss = match_train.groups()
epochs.append(int(epoch))
train_losses.append(float(loss))
if match_valid:
epoch, _, loss = match_valid.groups()
valid_epochs_loss.append(int(epoch))
valid_losses.append(float(loss))
if match_metric:
epoch, _, metric = match_metric.groups()
valid_epochs.append(int(epoch))
valid_metrics.append(float(metric))
return epochs, train_losses, valid_epochs_loss, valid_losses, valid_epochs, valid_metrics
# Extract data from the files
epochs1, train_losses1, valid_epochs_loss1, valid_losses1, valid_epochs1, valid_metrics1 = extract_data_from_file("./logs/training_log_2023-11-05_10:03:55.txt")
#epochs2, train_losses2, valid_epochs_loss2, valid_losses2, valid_epochs2, valid_metrics2 = extract_data_from_file("/Users/karanthakkar/Downloads/training_log_2023-11-04_21:34:50.txt")
#epochs3, train_losses3, valid_epochs_loss3, valid_losses3, valid_epochs3, valid_metrics3 = extract_data_from_file("/Users/karanthakkar/Downloads/training_log_2023-11-06_09:37:15.txt")
#epochs4, train_losses4, valid_epochs_loss4, valid_losses4, valid_epochs4, valid_metrics4 = extract_data_from_file("/Users/karanthakkar/Downloads/training_log_2023-11-05_23:11:26.txt")
# Plot Train Loss with log scale
plt.figure(figsize=(12, 6))
plt.plot(epochs1, train_losses1, '-o', label="Train Loss", markersize=3)
#plt.plot(epochs2, train_losses2, '-o', linestyle='--', label="Train Loss (random sampling across trials)", markersize=3)
#plt.plot(epochs3, train_losses3, '-o', linestyle='--', label="Train Loss (longer patience)", markersize=3)
#plt.plot(epochs4, train_losses4, '-o', linestyle='--', label="Train Loss (longer patience, different seed)", markersize=3)
plt.xlabel("Epoch")
plt.ylabel("Train Loss (Log Scale)")
#plt.yscale("log")
plt.title("Train Loss over Epochs")
plt.legend()
plt.grid(True, which="both", ls="--", linewidth=0.5)
plt.savefig("./logs/train_loss_overlayed_2.png")
# Continue with other plots...
# Plot Valid Loss
plt.figure(figsize=(12, 6))
plt.plot(valid_epochs_loss1, valid_losses1, '-o', label="Valid Loss", markersize=3)
#plt.plot(valid_epochs_loss2, valid_losses2, '-o', linestyle='--', label="Valid Loss (random sampling across trials)", markersize=3)
#plt.plot(valid_epochs_loss3, valid_losses3, '-o', linestyle='--', label="Valid Loss (longer patience)", markersize=3)
#plt.plot(valid_epochs_loss4, valid_losses4, '-o', linestyle='--', label="Valid Loss (longer patience, different seed)", markersize=3)
plt.xlabel("Epoch")
plt.ylabel("Valid Loss")
plt.title("Valid Loss over Epochs")
plt.legend()
plt.grid(True)
plt.savefig("./logs/valid_loss_overlayed_2.png")
# Plot Validation Metric
plt.figure(figsize=(12, 6))
plt.plot(valid_epochs1, valid_metrics1, '-o', label="Validation Metric", markersize=3)
#plt.plot(valid_epochs2, valid_metrics2, '-o', label="Validation Metric (random sampling across trials)", markersize=3, linestyle='--')
#plt.plot(valid_epochs3, valid_metrics3, '-o', label="Validation Metric (longer patience)", markersize=3, linestyle='--')
#plt.plot(valid_epochs4, valid_metrics4, '-o', label="Validation Metric (longer patience, different seed)", markersize=3, linestyle='--')
plt.xlabel("Epoch")
plt.ylabel("Metric Value")
plt.title("Validation Metric over Epochs")
plt.legend()
plt.grid(True)
plt.savefig("./logs/validation_metric_overlayed_2.png")