RocketGit

hahattpro / crowd_counting_framework (public) (License: Unspecified) (since 2020-12-23) (hash sha1)

backupfrom https://github.com/ttpro1995/crowd_counting_framework

Clone URLs: https://rocketgit.com/user/hahattpro/crowd_counting_framework ssh://rocketgit@ssh.rocketgit.com/user/hahattpro/crowd_counting_framework git://git.rocketgit.com/user/hahattpro/crowd_counting_framework

List of commits:

Subject	Hash	Author	Date (UTC)
add comet ml	a9d4b89ce594f5e241168ccafdcdf0f150ea0ebb	Thai Thien	2019-09-23 17:07:58
fix pacnn avg schema	c2140a96886195782e5689c24aeeb4fe7a2db7ad	Thai Thien	2019-09-22 17:35:01
debug number not divisible by 8	a568fd7f294a8bd31b3db78437b4b6b51b5b41b9	Thai Thien	2019-09-22 04:36:06
pacnn	967074890d14ab0eefc277801860270a468e8f9f	Thai Thien	2019-09-22 03:54:48
wip: pacnn	2192d7c7b449fecf3868877d9cfbc09bb6f7ae98	Thai Thien	2019-09-22 03:44:56
wip: pacnn	37620e5a9bc0f9516ea964ec58d9bdaa1c40ff36	Thai Thien	2019-09-22 03:14:42
fix training flow	2b87b1b26c7296b64493fdc49fedb421b249dfa3	Thai Thien	2019-09-17 18:00:35
dataset script	bc5c052f5f956510ab95ef9a45434fd486c57fae	Thai Thien	2019-09-16 17:21:13
evaluator	ffc5bf8290ae0c469a9a18a2d061cfd1bfeee822	Thai Thien	2019-09-14 04:56:35
some more test for data loader	25173578cde7d4e9fe6c6140d1ee01caa4fcfc32	Thai Thien	2019-09-14 02:51:58
some visualize to debug data loader	e4f52007616acf307bddbde79c0fb4f8c649c785	Thai Thien	2019-09-13 17:35:45
wip	d7d44cad6774355bdfa45414258763f6c6a0c299	Thai Thien	2019-08-31 16:58:16
commit all	6dad7a58f7dbf9fc288ce9dd3e92be538851c2a7	Thai Thien	2019-08-29 19:10:44
input d1,d2,d3 match	fc2a809241f8b6356d964c63d40cbebd55ca5f6c	Thai Thien	2019-08-28 17:57:05
WIP	39eab26d061e61dfffbf164dbd5fd878299b7250	thient	2019-08-28 11:09:12
output of de is ok	dd770386674df3e0fbebafdfc48a9352bc28967d	thient	2019-08-28 10:54:09
code pacnn	c49537b5cc91e96e4e35c9338d2c95b9bb41c672	Thai Thien	2019-08-27 16:35:27
crowd counting stuff	da9f27a39cba9bdd021b6b5c562f5f7c2be50190	Thai Thien	2019-08-24 18:27:44
seem ok	53fa176c31669a0e89b04adf290cb398f0316c45	Thai Thien	2019-08-24 18:26:31
flow ok	ad849681000818dfbcd0c1715c2858aed7236041	Thai Thien	2019-08-24 17:00:02

Commit a9d4b89ce594f5e241168ccafdcdf0f150ea0ebb - add comet ml
Author: Thai Thien
Author date (UTC): 2019-09-23 17:07
Committer name: Thai Thien
Committer date (UTC): 2019-09-23 17:07
Parent(s): c2140a96886195782e5689c24aeeb4fe7a2db7ad
Signing key:
Tree: 55eb23934aac7a2c7bc3bb97432ec9de64634781

File	Lines added	Lines deleted
README.md	2	0
main_pacnn.py	35	8
train_script/train_pacnn_shanghaitechA.sh	2	0

File README.md changed (mode: 100644) (index 7065abe..929964f)
3	3	conda create -n env python=3.7 anaconda	conda create -n env python=3.7 anaconda
4	4	conda install pytorch torchvision cpuonly -c pytorch	conda install pytorch torchvision cpuonly -c pytorch
5	5	conda install -c conda-forge opencv	conda install -c conda-forge opencv
	6		pip install comet_ml
	7
6	8	```	```

File main_pacnn.py changed (mode: 100644) (index 9024f16..a21c0f5)
	1		from comet_ml import Experiment
1	2	from args_util import real_args_parse	from args_util import real_args_parse
2	3	from data_flow import get_train_val_list, get_dataloader, create_training_image_list	from data_flow import get_train_val_list, get_dataloader, create_training_image_list
3	4	from ignite.engine import Events, create_supervised_trainer, create_supervised_evaluator	from ignite.engine import Events, create_supervised_trainer, create_supervised_evaluator

...	...	from evaluator import MAECalculator
18	19	from model_util import save_checkpoint	from model_util import save_checkpoint
19	20
20	21	if __name__ == "__main__":	if __name__ == "__main__":
	22		# import comet_ml in the top of your file
	23
	24
	25		MODEL_SAVE_NAME = "dev4"
	26		# Add the following code anywhere in your machine learning file
	27		experiment = Experiment(api_key="S3mM1eMq6NumMxk2QJAXASkUM",
	28		project_name="pacnn-dev", workspace="ttpro1995")
	29		experiment.set_name(MODEL_SAVE_NAME)
	30
21	31	device = torch.device("cuda" if torch.cuda.is_available() else "cpu")	device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
22	32
23	33	# device = "cpu"	# device = "cpu"

...	...	if __name__ == "__main__":
28	38	DATASET_NAME = "shanghaitech"	DATASET_NAME = "shanghaitech"
29	39	PACNN_PERSPECTIVE_AWARE_MODEL = False	PACNN_PERSPECTIVE_AWARE_MODEL = False
30	40
	41
	42
31	43	# create list	# create list
32	44	if DATASET_NAME is "shanghaitech":	if DATASET_NAME is "shanghaitech":
33	45	TRAIN_PATH = os.path.join(DATA_PATH, "train_data")	TRAIN_PATH = os.path.join(DATA_PATH, "train_data")

...	...	if __name__ == "__main__":
72	84	optimizer = torch.optim.SGD(net.parameters(), args.lr,	optimizer = torch.optim.SGD(net.parameters(), args.lr,
73	85	momentum=args.momentum,	momentum=args.momentum,
74	86	weight_decay=args.decay)	weight_decay=args.decay)
75		for e in range(1):
	87		for e in range(10):
76	88	print("start epoch ", e)	print("start epoch ", e)
77	89	loss_sum = 0	loss_sum = 0
78	90	sample = 0	sample = 0

...	...	if __name__ == "__main__":
107	119	sample += 1	sample += 1
108	120	optimizer.zero_grad()	optimizer.zero_grad()
109	121	counting += 1	counting += 1
	122
110	123	if counting%10 ==0:	if counting%10 ==0:
111		print("counting ", counting, " -- avg loss", loss_sum/sample)
	124		avg_loss_ministep = loss_sum/sample
	125		print("counting ", counting, " -- avg loss ", avg_loss_ministep)
	126		experiment.log_metric("avg_loss_ministep", avg_loss_ministep)
112	127	# if counting == 100:	# if counting == 100:
113	128	# break	# break
114
115	129	end_time = time()	end_time = time()
116	130	avg_loss = loss_sum/sample	avg_loss = loss_sum/sample
117	131	epoch_time = end_time - start_time	epoch_time = end_time - start_time
	132		print("==END epoch ", e, " =============================================")
118	133	print(epoch_time, avg_loss, sample)	print(epoch_time, avg_loss, sample)
119
	134		experiment.log_metric("avg_loss_epoch", avg_loss)
	135		print("=================================================================")
120	136
121	137	save_checkpoint({	save_checkpoint({
122	138	'state_dict': net.state_dict(),	'state_dict': net.state_dict(),
123		}, False, "test2")
	139		}, False, MODEL_SAVE_NAME)
124	140
125	141
126	142

...	...	if __name__ == "__main__":
130	146	net = PACNNWithPerspectiveMap(PACNN_PERSPECTIVE_AWARE_MODEL).to(device)	net = PACNNWithPerspectiveMap(PACNN_PERSPECTIVE_AWARE_MODEL).to(device)
131	147	print(net)	print(net)
132	148
133		# best_checkpoint = torch.load("test2checkpoint.pth.tar")
134		# net.load_state_dict(best_checkpoint['state_dict'])
	149		best_checkpoint = torch.load(MODEL_SAVE_NAME + "checkpoint.pth.tar")
	150		net.load_state_dict(best_checkpoint['state_dict'])
135	151
136	152	# device = "cpu"	# device = "cpu"
	153		# TODO d1_val 155.97279205322266
	154		# d2_val 35.46327234903971
	155		# d3_val 23.07176342010498
	156		# why d2 and d3 mse too low
137	157	mae_calculator_d1 = MAECalculator()	mae_calculator_d1 = MAECalculator()
138	158	mae_calculator_d2 = MAECalculator()	mae_calculator_d2 = MAECalculator()
139	159	mae_calculator_d3 = MAECalculator()	mae_calculator_d3 = MAECalculator()
	160		mae_calculator_final = MAECalculator()
140	161	with torch.no_grad():	with torch.no_grad():
141	162	for val_img, label in val_loader_pacnn:	for val_img, label in val_loader_pacnn:
142	163	net.eval()	net.eval()

...	...	if __name__ == "__main__":
144	165	d1_label, d2_label, d3_label = label	d1_label, d2_label, d3_label = label
145	166
146	167	# forward pass	# forward pass
147		d1, d2, d3 = net(val_img.to(device))
	168		d1, d2, d3, p_s, p, d = net(val_img.to(device))
148	169
149	170	d1_label = d1_label.to(device)	d1_label = d1_label.to(device)
150	171	d2_label = d2_label.to(device)	d2_label = d2_label.to(device)

...	...	if __name__ == "__main__":
154	175	mae_calculator_d1.eval(d1.cpu().detach().numpy(), d1_label.cpu().detach().numpy())	mae_calculator_d1.eval(d1.cpu().detach().numpy(), d1_label.cpu().detach().numpy())
155	176	mae_calculator_d2.eval(d2.cpu().detach().numpy(), d2_label.cpu().detach().numpy())	mae_calculator_d2.eval(d2.cpu().detach().numpy(), d2_label.cpu().detach().numpy())
156	177	mae_calculator_d3.eval(d3.cpu().detach().numpy(), d3_label.cpu().detach().numpy())	mae_calculator_d3.eval(d3.cpu().detach().numpy(), d3_label.cpu().detach().numpy())
	178		mae_calculator_final.eval(d.cpu().detach().numpy(), d1_label.cpu().detach().numpy())
157	179	print("count ", mae_calculator_d1.count)	print("count ", mae_calculator_d1.count)
158	180	print("d1_val ", mae_calculator_d1.get_mae())	print("d1_val ", mae_calculator_d1.get_mae())
159	181	print("d2_val ", mae_calculator_d2.get_mae())	print("d2_val ", mae_calculator_d2.get_mae())
160	182	print("d3_val ", mae_calculator_d3.get_mae())	print("d3_val ", mae_calculator_d3.get_mae())
	183		print("dfinal_val ", mae_calculator_final.get_mae())
	184		experiment.log_metric("d1_val", mae_calculator_d1.get_mae())
	185		experiment.log_metric("d2_val", mae_calculator_d2.get_mae())
	186		experiment.log_metric("d3_val", mae_calculator_d3.get_mae())
	187		experiment.log_metric("dfinal_val", mae_calculator_final.get_mae())
161	188
162	189

File train_script/train_pacnn_shanghaitechA.sh changed (mode: 100644) (index d90895d..34be054)
	1		#python /home/tt/project/crowd_counting_framework/main_pacnn.py --input /home/tt/project/crowd_counting_framework/data/ShanghaiTech/part_A
	2
1	3	python /home/tt/project/crowd_counting_framework/main_pacnn.py --input /home/tt/project/crowd_counting_framework/data/ShanghaiTech/part_A	python /home/tt/project/crowd_counting_framework/main_pacnn.py --input /home/tt/project/crowd_counting_framework/data/ShanghaiTech/part_A

Hints:
Before first commit, do not forget to setup your git environment:

git config --global user.name "your_name_here"
git config --global user.email "your@email_here"

Clone this repository using HTTP(S):

git clone https://rocketgit.com/user/hahattpro/crowd_counting_framework

Clone this repository using ssh (do not forget to upload a key first):

git clone ssh://rocketgit@ssh.rocketgit.com/user/hahattpro/crowd_counting_framework

Clone this repository using git:

git clone git://git.rocketgit.com/user/hahattpro/crowd_counting_framework

You are allowed to anonymously push to this repository.
This means that your pushed commits will automatically be transformed into a merge request:

... clone the repository ...
... make some changes and some commits ...
git push origin main