-
Notifications
You must be signed in to change notification settings - Fork 60
/
train_densify_prune.py
executable file
·281 lines (248 loc) · 9.75 KB
/
train_densify_prune.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
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
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
#
# This software is free for non-commercial, research and evaluation use
# under the terms of the LICENSE.md file.
#
# For inquiries contact [email protected]
#
import os
import torch
from random import randint
from utils.loss_utils import l1_loss, ssim
from gaussian_renderer import render, network_gui
import sys
from lpipsPyTorch import lpips
from scene import Scene, GaussianModel
from utils.general_utils import safe_state
from utils.logger_utils import training_report, prepare_output_and_logger
import uuid
from tqdm import tqdm
from utils.image_utils import psnr
from argparse import ArgumentParser, Namespace
from arguments import ModelParams, PipelineParams, OptimizationParams
# from prune_train import prepare_output_and_logger, training_report
from icecream import ic
from os import makedirs
from prune import prune_list, calculate_v_imp_score
import torchvision
from torch.optim.lr_scheduler import ExponentialLR
import csv
import numpy as np
try:
from torch.utils.tensorboard import SummaryWriter
TENSORBOARD_FOUND = True
except ImportError:
TENSORBOARD_FOUND = False
def training(
dataset,
opt,
pipe,
testing_iterations,
saving_iterations,
checkpoint_iterations,
checkpoint,
debug_from,
args,
):
first_iter = 0
tb_writer = prepare_output_and_logger(dataset)
gaussians = GaussianModel(dataset.sh_degree)
scene = Scene(dataset, gaussians)
gaussians.training_setup(opt)
if checkpoint:
(model_params, first_iter) = torch.load(checkpoint)
gaussians.restore(model_params, opt)
bg_color = [1, 1, 1] if dataset.white_background else [0, 0, 0]
background = torch.tensor(bg_color, dtype=torch.float32, device="cuda")
iter_start = torch.cuda.Event(enable_timing=True)
iter_end = torch.cuda.Event(enable_timing=True)
viewpoint_stack = None
ema_loss_for_log = 0.0
progress_bar = tqdm(range(first_iter, opt.iterations), desc="Training progress")
first_iter += 1
gaussians.scheduler = ExponentialLR(gaussians.optimizer, gamma=0.97)
for iteration in range(first_iter, opt.iterations + 1):
if network_gui.conn == None:
network_gui.try_connect()
while network_gui.conn != None:
try:
net_image_bytes = None
(
custom_cam,
do_training,
pipe.convert_SHs_python,
pipe.compute_cov3D_python,
keep_alive,
scaling_modifer,
) = network_gui.receive()
if custom_cam != None:
net_image = render(
custom_cam, gaussians, pipe, background, scaling_modifer
)["render"]
net_image_bytes = memoryview(
(torch.clamp(net_image, min=0, max=1.0) * 255)
.byte()
.permute(1, 2, 0)
.contiguous()
.cpu()
.numpy()
)
network_gui.send(net_image_bytes, dataset.source_path)
if do_training and (
(iteration < int(opt.iterations)) or not keep_alive
):
break
except Exception as e:
network_gui.conn = None
iter_start.record()
gaussians.update_learning_rate(iteration)
# Every 1000 its we increase the levels of SH up to a maximum degree
if iteration % 1000 == 0:
gaussians.oneupSHdegree()
gaussians.scheduler.step()
# Pick a random Camera
if not viewpoint_stack:
viewpoint_stack = scene.getTrainCameras().copy()
viewpoint_cam = viewpoint_stack.pop(randint(0, len(viewpoint_stack) - 1))
# Render
if (iteration - 1) == debug_from:
pipe.debug = True
render_pkg = render(viewpoint_cam, gaussians, pipe, background)
image, viewspace_point_tensor, visibility_filter, radii = (
render_pkg["render"],
render_pkg["viewspace_points"],
render_pkg["visibility_filter"],
render_pkg["radii"],
)
# Loss
gt_image = viewpoint_cam.original_image.cuda()
Ll1 = l1_loss(image, gt_image)
loss = (1.0 - opt.lambda_dssim) * Ll1 + opt.lambda_dssim * (
1.0 - ssim(image, gt_image)
)
loss.backward()
iter_end.record()
with torch.no_grad():
# Progress bar
ema_loss_for_log = 0.4 * loss.item() + 0.6 * ema_loss_for_log
if iteration % 10 == 0:
progress_bar.set_postfix({"Loss": f"{ema_loss_for_log:.{7}f}"})
progress_bar.update(10)
if iteration == opt.iterations:
progress_bar.close()
# Log and save
if iteration in saving_iterations:
print("\n[ITER {}] Saving Gaussians".format(iteration))
scene.save(iteration)
training_report(
tb_writer,
iteration,
Ll1,
loss,
l1_loss,
iter_start.elapsed_time(iter_end),
testing_iterations,
scene,
render,
(pipe, background),
)
# Densification
if iteration < opt.densify_until_iter:
# Keep track of max radii in image-space for pruning
gaussians.max_radii2D[visibility_filter] = torch.max(
gaussians.max_radii2D[visibility_filter], radii[visibility_filter]
)
gaussians.add_densification_stats(
viewspace_point_tensor, visibility_filter
)
if (
iteration > opt.densify_from_iter
and iteration % opt.densification_interval == 0
):
size_threshold = (
20 if iteration > opt.opacity_reset_interval else None
)
gaussians.densify_and_prune(
opt.densify_grad_threshold,
0.005,
scene.cameras_extent,
size_threshold,
)
if iteration % opt.opacity_reset_interval == 0 or (
dataset.white_background and iteration == opt.densify_from_iter
):
gaussians.reset_opacity()
if iteration in args.prune_iterations:
# TODO Add prunning types
gaussian_list, imp_list = prune_list(gaussians, scene, pipe, background)
i = args.prune_iterations.index(iteration)
v_list = calculate_v_imp_score(gaussians, imp_list, args.v_pow)
gaussians.prune_gaussians(
(args.prune_decay**i) * args.prune_percent, v_list
)
# Optimizer step
if iteration < opt.iterations:
gaussians.optimizer.step()
gaussians.optimizer.zero_grad(set_to_none=True)
if iteration in checkpoint_iterations:
print("\n[ITER {}] Saving Checkpoint".format(iteration))
if not os.path.exists(scene.model_path):
os.makedirs(scene.model_path)
torch.save(
(gaussians.capture(), iteration),
scene.model_path + "/chkpnt" + str(iteration) + ".pth",
)
if iteration == checkpoint_iterations[-1]:
gaussian_list, imp_list = prune_list(gaussians, scene, pipe, background)
v_list = calculate_v_imp_score(gaussians, imp_list, args.v_pow)
np.savez(os.path.join(scene.model_path,"imp_score"), v_list.cpu().detach().numpy())
if __name__ == "__main__":
# Set up command line argument parser
parser = ArgumentParser(description="Training script parameters")
lp = ModelParams(parser)
op = OptimizationParams(parser)
pp = PipelineParams(parser)
parser.add_argument("--ip", type=str, default="127.0.0.1")
parser.add_argument("--port", type=int, default=6009)
parser.add_argument("--debug_from", type=int, default=-1)
parser.add_argument("--detect_anomaly", action="store_true", default=False)
parser.add_argument(
"--test_iterations",
nargs="+",
type=int,
default=[7_000, 30_000],
)
parser.add_argument(
"--save_iterations", nargs="+", type=int, default=[7_000, 30_000]
)
parser.add_argument("--quiet", action="store_true")
parser.add_argument(
"--checkpoint_iterations", nargs="+", type=int, default=[7_000, 30_000]
)
parser.add_argument("--start_checkpoint", type=str, default=None)
parser.add_argument(
"--prune_iterations", nargs="+", type=int, default=[16_000, 24_000]
)
parser.add_argument("--prune_percent", type=float, default=0.5)
parser.add_argument("--v_pow", type=float, default=0.1)
parser.add_argument("--prune_decay", type=float, default=0.8)
args = parser.parse_args(sys.argv[1:])
args.save_iterations.append(args.iterations)
print("Optimizing " + args.model_path)
# Initialize system state (RNG)
safe_state(args.quiet)
# Start GUI server, configure and run training
network_gui.init(args.ip, args.port)
torch.autograd.set_detect_anomaly(args.detect_anomaly)
training(
lp.extract(args),
op.extract(args),
pp.extract(args),
args.test_iterations,
args.save_iterations,
args.checkpoint_iterations,
args.start_checkpoint,
args.debug_from,
args,
)
# All done
print("\nTraining complete.")