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
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
|
import torch
import numpy as np
import torch.optim as optim
from tqdm import tqdm
from torch.utils.data import DataLoader
import torchaudio
import torch.nn.functional as F
from neurobench.datasets import SpeechCommands
from ANN import M5
BATCH_SIZE = 256
NUM_WORKERS = 8
EPOCHS = 50
# Check if GPU is available
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
print(device)
# Load the dataset
data_dir = "../../../data/speech_commands/"
train_set = SpeechCommands(path=data_dir, subset="training")
val_set = SpeechCommands(path=data_dir, subset="validation")
test_set = SpeechCommands(path=data_dir, subset="testing")
# Create the dataloaders
train_loader = DataLoader(train_set, batch_size=BATCH_SIZE, shuffle=True)
val_loader = DataLoader(val_set, batch_size=BATCH_SIZE, shuffle=True)
test_loader = DataLoader(test_set, batch_size=BATCH_SIZE, shuffle=True)
# Model
model = M5()
model.to(device)
transform = torchaudio.transforms.Resample(orig_freq=16000, new_freq=8000)
optimizer = optim.Adam(model.parameters(), lr=0.01, weight_decay=0.0001)
scheduler = optim.lr_scheduler.StepLR(optimizer, step_size=20, gamma=0.1)
def get_likely_index(tensor):
# find most likely label index for each element in the batch
return tensor.argmax(dim=-1)
def number_of_correct(pred, target):
# count number of correct predictions
return pred.squeeze().eq(target).sum().item()
def train(model, epoch, log_interval):
model.train()
for batch_idx, (data, target) in enumerate(train_loader):
data = data.permute(0, 2, 1).to(device)
target = target.to(device)
# apply transform and model on whole batch directly on device
data = transform(data)
output = model(data)
# negative log-likelihood for a tensor of size (batch x 1 x n_output)
loss = F.nll_loss(output.squeeze(), target)
optimizer.zero_grad()
loss.backward()
optimizer.step()
# print training stats
if batch_idx % log_interval == 0:
print(f"Train Epoch: {epoch} [{batch_idx * len(data)}/{len(train_loader.dataset)} ({100. * batch_idx / len(train_loader):.0f}%)]\tLoss: {loss.item():.6f}")
def validate(model, epoch):
model.eval()
correct = 0
for batch_idx, (data, target) in enumerate(val_loader):
data = data.permute(0, 2, 1).to(device)
target = target.to(device)
# apply transform and model on whole batch directly on device
data = transform(data)
output = model(data)
pred = get_likely_index(output)
correct += number_of_correct(pred, target)
return correct / len(val_loader.dataset)
# Training Start
best_acc = 0
for epoch in range(EPOCHS):
print(f"Epoch {epoch}:")
model.train()
train(model, epoch, log_interval=20)
val_acc = []
# validate
val_acc.append(validate(model, epoch))
print(f"Validation Accuracy: {np.mean(val_acc) * 100:.2f}%")
if np.mean(val_acc) > best_acc:
print("New Best Validation Accuracy. Saving...")
best_acc = np.mean(val_acc)
torch.save(model.state_dict(), "model_data/m5_ann")
scheduler.step()
print(f"---------------------\n")
# Load the weights into the network for inference
model.load_state_dict(torch.load("model_data/m5_ann"))
|
