Diffusion_Classifier/train.py at master · AniketGurav/Diffusion_Classifier · GitHub

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
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
import os
import argparse
from tqdm import tqdm
import torch
from torch.utils.data import DataLoader
from torchvision import transforms
from torchvision.datasets import MNIST
from torchvision.utils import save_image, make_grid
import matplotlib.pyplot as plt
from matplotlib.animation import FuncAnimation, PillowWriter
from model import DDPM, ContextUnet

def get_args_parser():
    parser = argparse.ArgumentParser('Guided Diffusion', add_help=False)
    parser.add_argument('--batch_size', default=512, type=int)
    parser.add_argument('--epochs', default=20, type=int)

    # Model parameters
    parser.add_argument('--n_feat', default=256, type=int)
    parser.add_argument('--n_T', default=400, type=int)

    # Optimizer parameters
    parser.add_argument('--lr', type=float, default=1e-4, metavar='LR',
                        help='learning rate (absolute lr)')

    # Dataset parameters
    parser.add_argument('--data_path', default='./data', type=str,
                        help='dataset path')
    parser.add_argument('--n_classes', default=10, type=int,
                        help='number of the classification types')

    parser.add_argument('--output_dir', default='./out/',
                        help='path where to save, empty for no saving')
    parser.add_argument('--device', default='cuda',
                        help='device to use for training / testing')
    parser.add_argument('--seed', default=0, type=int)

    parser.add_argument('--start_epoch', default=0, type=int, metavar='N',
                        help='start epoch')
    parser.add_argument('--num_workers', default=6, type=int)
    parser.add_argument('--eval', action='store_true')
    parser.add_argument('--save_model', action='store_true')

    return parser


def train_mnist(args):

    ddpm = DDPM(nn_model=ContextUnet(in_channels=1, n_feat=args.n_feat, n_classes=args.n_classes),
                n_classes=args.n_classes, betas=(1e-4, 0.02), n_T=args.n_T, device=args.device, drop_prob=0.1)
    ddpm.to(args.device)

    print('Model = %s' % str(ddpm))

    # optionally load a model
    # ddpm.load_state_dict(torch.load('./data/diffusion_outputs/ddpm_unet01_mnist_9.pth'))

    tf = transforms.Compose([transforms.ToTensor()]) # mnist is already normalised 0 to 1
    dataset = MNIST(args.data_path, train=True, download=True, transform=tf)
    dataloader = DataLoader(dataset, batch_size=args.batch_size, shuffle=True, num_workers=8)
    optim = torch.optim.Adam(ddpm.parameters(), lr=args.lr)

    eval_freq = 5
    for ep in range(args.start_epoch, args.epochs):
        print(f'epoch {ep}')
        ddpm.train()

        # linear lrate decay
        optim.param_groups[0]['lr'] = args.lr*(1-ep/args.epochs)

        pbar = tqdm(dataloader)
        loss_ema = None
        for x, c in pbar:
            optim.zero_grad()
            x = x.to(args.device)
            c = c.to(args.device)
            loss = ddpm.loss(x, c)
            loss.backward()
            if loss_ema is None:
                loss_ema = loss.item()
            else:
                loss_ema = 0.95 * loss_ema + 0.05 * loss.item()
            pbar.set_description(f'loss: {loss_ema:.4f}')
            optim.step()


        if args.eval and (ep%eval_freq==0 or ep == int(args.epochs-1)):
            # for eval, save an image of currently generated samples (top rows)
            # followed by real images (bottom rows)
            ddpm.eval()
            ws_test = [0.0, 2.0] # strength of generative guidance
            with torch.no_grad():
                n_sample = 4*args.n_classes
                for w_i, w in enumerate(ws_test):
                    x_gen, x_gen_store = ddpm.sample(n_sample, (1, 28, 28), args.device, guide_w=w)

                    # append some real images at bottom, order by class also
                    x_real = torch.Tensor(x_gen.shape).to(args.device)
                    for k in range(args.n_classes):
                        for j in range(int(n_sample/args.n_classes)):
                            try:
                                idx = torch.squeeze((c == k).nonzero())[j]
                            except:
                                idx = 0
                            x_real[k+(j*args.n_classes)] = x[idx]

                    x_all = torch.cat([x_gen, x_real])
                    grid = make_grid(x_all*-1 + 1, nrow=10)
                    image_path = os.path.join(args.output_dir, f'image_ep{ep}_w{w}.png')
                    save_image(grid, image_path)
                    print('saved image at ' + image_path)

                    # create gif of images evolving over time, based on x_gen_store
                    fig, axs = plt.subplots(nrows=int(n_sample/args.n_classes), ncols=args.n_classes,
                                            sharex=True, sharey=True, figsize=(8,3))
                    def animate_diff(i, x_gen_store):
                        print(f'gif animating frame {i} of {x_gen_store.shape[0]}', end='\r')
                        plots = []
                        for row in range(int(n_sample/args.n_classes)):
                            for col in range(args.n_classes):
                                axs[row, col].clear()
                                axs[row, col].set_xticks([])
                                axs[row, col].set_yticks([])
                                plots.append(axs[row, col].imshow(-x_gen_store[i,(row*args.n_classes)+col,0],
                                                                  cmap='gray',
                                                                  vmin=(-x_gen_store[i]).min(),
                                                                  vmax=(-x_gen_store[i]).max()))
                        return plots
                    ani = FuncAnimation(fig, animate_diff, fargs=[x_gen_store],
                                        interval=200, blit=False, repeat=True, frames=x_gen_store.shape[0])
                    gif_path = os.path.join(args.output_dir, f'gif_ep{ep}_w{w}.gif')
                    ani.save(gif_path, dpi=100, writer=PillowWriter(fps=5))
                    print('saved image at ' + gif_path)

        # optionally save model
        if args.save_model and ep == int(args.epochs-1):
            model_path = os.path.join(args.output_dir, f'model_{ep}.pth')
            torch.save(ddpm.state_dict(), model_path)
            print('saved model at ' + model_path)

if __name__ == '__main__':
    args = get_args_parser()
    args = args.parse_args()
    train_mnist(args)