MuST/train.py at main · MrCrims/MuST · 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
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
import os
import time
import torch
import numpy as np
from options import *
from model.fwm import FWM

import utils.dataloader as dataloader
import utils.utils as utils
import logging
from collections import defaultdict
from average_meter import AverageMeter
import torchvision

def train(model:FWM,
          device: torch.device,
          fwm_config: FWM_Options,
          train_options: TrainingOptions,
          this_run_folder: str,
          writer):
    """
    Trains the FWM model
    :param model: The model
    :param device: torch.device object, usually this is GPU (if avaliable), otherwise CPU.
    :param fwm_config: The network configuration
    :param train_options: The training settings
    :param this_run_folder: The parent folder for the current training run to store training artifacts/results/logs.
    :param tb_logger: TensorBoardLogger object which is a thin wrapper for TensorboardX logger.
                Pass None to disable TensorboardX logging
    :return:
    """
    bg_loader = dataloader.dataloader_background(train_options.background_folder,1)# default batchsize of  background 1
    train_loader = dataloader.dataloader_img_mask(train_options.train_folder,train_options.batch_size)
    l = len(train_loader)
    rate = 0.05
    val_l = int(l*rate)
    train_l = l - val_l
    # val_laoder = dataloader.dataloader_img_mask(train_options.validation_folder,train_options.batch_size)
    file_count = len(train_loader.dataset)
    if file_count % train_options.batch_size == 0:
        steps_in_epoch = file_count // train_options.batch_size
    else:
        steps_in_epoch = file_count // train_options.batch_size + 1


    print_each = 10

    for epoch in range(train_options.start_epoch,train_options.number_of_epochs+1):
        logging.info('\nStarting epoch {}/{}'.format(epoch, train_options.number_of_epochs))
        logging.info('Batch size = {}\nSteps in epoch = {}'.format(train_options.batch_size, steps_in_epoch))
        training_losses = defaultdict(AverageMeter)
        epoch_start = time.time()
        step = 1
        for i in range(train_l):
            images,masks = train_loader.__iter__().__next__()
            images = images.to(device)
            masks = masks.to(device)
            masks[masks>0]=1
            masks[masks<1]=0
            messages = torch.Tensor(np.random.choice([0, 1], (images.shape[0], fwm_config.message_length))).to(device)
            background = bg_loader.__iter__().__next__()
            background = background.to(device)
            losses , (merge_img , pred_mask ,final_mask) = model.train_on_batch([background,images,masks,messages],epoch)
            # if not torch.is_tensor(merge_img):
            #     continue
            for name , loss in losses.items():
                training_losses[name].update(loss)
            if (step % print_each == 0 or step == steps_in_epoch) and writer is not None:
                logging.info(
                    'Epoch: {}/{} Step: {}/{}'.format(epoch, train_options.number_of_epochs, step, steps_in_epoch))
                utils.log_progress(training_losses)
                logging.info('-' * 40)
                writer.add_scalar("train loss",losses['loss           '],epoch*(len(train_loader))+step)


            step += 1
        model.lr_scheduler.step()
        # model.lr_scheduler_enc.step()
        train_duration = time.time() - epoch_start
        logging.info('Epoch {} training duration {:.2f} sec'.format(epoch, train_duration))
        logging.info('-' * 40)
        utils.write_losses(os.path.join(this_run_folder, 'train.csv'), training_losses, epoch, train_duration)

        validation_losses = defaultdict(AverageMeter)
        logging.info('Running validation for epoch {}/{}'.format(epoch, train_options.number_of_epochs))
        step = 1
        for i in range(val_l):
            images,masks = train_loader.__iter__().__next__()
            images = images.to(device)
            masks = masks.to(device)
            masks[masks>0]=1
            masks[masks<1]=0
            if writer is not None:
                grid = torchvision.utils.make_grid(images)
                writer.add_image("batch of image",grid)
                grid = torchvision.utils.make_grid(masks)
                writer.add_image("masks",grid)
            messages = torch.Tensor(np.random.choice([0, 1], (images.shape[0], fwm_config.message_length))).to(device)
            background = bg_loader.__iter__().__next__()
            background = background.to(device)

            losses , (merge_img , pred_mask ,final_mask) = model.validation_on_batch([background,images,masks,messages])
            # if not torch.is_tensor(merge_img):
            #     continue
            for name , loss in losses.items():
                validation_losses[name].update(loss)
            if (step % print_each == 0 or step == steps_in_epoch) and writer is not None:
                writer.add_scalar("validation loss",losses['dec_mse        '],epoch*(len(train_loader))+step)
            if step%100==0 and writer is not None:
                grid = torchvision.utils.make_grid(merge_img)
                writer.add_image("merge_img",grid)
                tobe_save = torch.cat([pred_mask,final_mask],dim=0)
                grid = torchvision.utils.make_grid(tobe_save)
                writer.add_image("pred_mask",grid)
            step += 1
        utils.log_progress(validation_losses)
        logging.info('-' * 40)
        utils.save_checkpoint(model, train_options.experiment_name, epoch, os.path.join(this_run_folder, 'checkpoints'))
        utils.write_losses(os.path.join(this_run_folder, 'validation.csv'), validation_losses, epoch,
                           time.time() - epoch_start)


def train_loc(model:FWM,
          device: torch.device,
          fwm_config: FWM_Options,
          train_options: TrainingOptions,
          this_run_folder: str,
          writer):
    """
    Trains the FWM model
    :param model: The model
    :param device: torch.device object, usually this is GPU (if avaliable), otherwise CPU.
    :param fwm_config: The network configuration
    :param train_options: The training settings
    :param this_run_folder: The parent folder for the current training run to store training artifacts/results/logs.
    :param tb_logger: TensorBoardLogger object which is a thin wrapper for TensorboardX logger.
                Pass None to disable TensorboardX logging
    :return:
    """
    bg_loader = dataloader.dataloader_background(train_options.background_folder,1)# default batchsize of  background 1
    train_loader = dataloader.dataloader_img_mask(train_options.train_folder,train_options.batch_size)
    l = len(train_loader)
    rate = 0.1
    val_l = int(l*rate)
    train_l = l - val_l
    # val_laoder = dataloader.dataloader_img_mask(train_options.validation_folder,train_options.batch_size)
    file_count = len(train_loader.dataset)
    if file_count % train_options.batch_size == 0:
        steps_in_epoch = file_count // train_options.batch_size
    else:
        steps_in_epoch = file_count // train_options.batch_size + 1


    print_each = 10

    for epoch in range(train_options.start_epoch,train_options.number_of_epochs+1):
        logging.info('\nStarting epoch {}/{}'.format(epoch, train_options.number_of_epochs))
        logging.info('Batch size = {}\nSteps in epoch = {}'.format(train_options.batch_size, steps_in_epoch))
        training_losses = defaultdict(AverageMeter)
        epoch_start = time.time()
        step = 1
        for i in range(train_l):
            images,masks = train_loader.__iter__().__next__()
            images = images.to(device)
            masks = masks.to(device)

            messages = torch.Tensor(np.random.choice([0, 1], (images.shape[0], fwm_config.message_length))).to(device)
            background = bg_loader.__iter__().__next__()
            background = background.to(device)
            losses , (merge_img) = model.train_on_batch_loc([background,images,masks,messages],step)

            for name , loss in losses.items():
                training_losses[name].update(loss)
            if (step % print_each == 0 or step == steps_in_epoch) and writer is not None:
                logging.info(
                    'Epoch: {}/{} Step: {}/{}'.format(epoch, train_options.number_of_epochs, step, steps_in_epoch))
                utils.log_progress(training_losses)
                logging.info('-' * 40)
                writer.add_scalar("train loss",losses['loss           '],epoch*(len(train_loader))+step)


            step += 1

        train_duration = time.time() - epoch_start
        logging.info('Epoch {} training duration {:.2f} sec'.format(epoch, train_duration))
        logging.info('-' * 40)
        utils.write_losses(os.path.join(this_run_folder, 'train.csv'), training_losses, epoch, train_duration)

        validation_losses = defaultdict(AverageMeter)
        logging.info('Running validation for epoch {}/{}'.format(epoch, train_options.number_of_epochs))
        step = 1
        for j in range(val_l):
            images,masks = train_loader.__iter__().__next__()
            images = images.to(device)
            masks = masks.to(device)
            # grid = torchvision.utils.make_grid(images)
            # writer.add_image("batch of image",grid)
            # grid = torchvision.utils.make_grid(masks)
            # writer.add_image("masks",grid)
            messages = torch.Tensor(np.random.choice([0, 1], (images.shape[0], fwm_config.message_length))).to(device)
            background = bg_loader.__iter__().__next__()
            background = background.to(device)

            losses , (merge_img) = model.validation_on_batch_loc([background,images,masks,messages])
            for name , loss in losses.items():
                validation_losses[name].update(loss)
            if (step % print_each == 0 or step == steps_in_epoch) and writer is not None:
                writer.add_scalar("validation loss",losses['loss           '],epoch*(len(train_loader))+step)
            # if step%100==0:
            #     grid = torchvision.utils.make_grid(merge_img)
            #     writer.add_image("merge_img",grid)
            #     tobe_save = torch.cat([pred_mask,final_mask],dim=0)
            #     grid = torchvision.utils.make_grid(tobe_save)
            #     writer.add_image("pred_mask",grid)
            step += 1
        utils.log_progress(validation_losses)
        logging.info('-' * 40)
        utils.save_checkpoint(model, train_options.experiment_name, epoch, os.path.join(this_run_folder, 'checkpoints'))
        utils.write_losses(os.path.join(this_run_folder, 'validation.csv'), validation_losses, epoch,
                           time.time() - epoch_start)