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
debug/a-1 develop feature/add_can_adcrowdnet feature/add_context_aware_network feature/attn_can_adcrowd_simple feature/cache_dataloader feature/compact_cnn feature/compare_with_original_count feature/dataloader_not_divide_8 feature/debug_viz feature/documentation feature/improve_density_gen feature/improve_loss_fn feature/loader_full_and_augment feature/refactor-dataflow feature/refactor_ccnn_basecode feature/save_best feature/simple_cometml feature/train_context_aware_network_on_shtA feature/train_dev_test_split hotfix/fix_visualize_data_loader master meow_exp server-training with_apex
List of commits:
Subject Hash Author Date (UTC)
no more batch timer 3fea48c4642f62ab92accbe2493b6275d68c1b02 Thai Thien 2020-04-30 19:04:16
num_worker=0 9e21a25fd4194e99c7899f7c6b6f04d51b432c61 Thai Thien 2020-04-30 19:01:09
.to(device) 136bafc672194946e944e85d43880f3342abf9be Thai Thien 2020-04-30 18:52:05
we load, so increase epoch b147d3d824c0597390900cb3506495ad9d842bd4 Thai Thien 2020-04-30 18:45:20
wrong note 31e67446689b4cecb8f118c43e6794fa437e8eb5 Thai Thien 2020-04-30 18:44:38
use ssim ccnn 311308b98fc0a426948953ecdf8b9eec6dfd69dd Thai Thien 2020-04-30 18:38:25
loss reduce as we train, it might ok cf1b359106f6e707e6862f2b3cbc9b5c85466ea4 Thai Thien 2020-04-30 18:34:19
replace SSIM with kornia SSIM 1f29782fe73751182b696c18567c1de3ab84a6bc Thai Thien 2020-04-30 17:13:09
fix link 3c55cee5cbc9fbaee0859b6ab61529029caee681 Thai Thien 2020-04-30 15:41:10
update documentation feaaa4095d89e99f71f0684eec302d01e421708d Thai Thien 2020-04-30 15:35:38
change gpu c7c6aa29ccee5736c2d4cb290c2ff71c6d897d57 Thai Thien 2020-04-30 12:31:11
reduce batch size 2b2820095211e117c6d02236ef85adb89b70098f Thai Thien 2020-04-30 12:28:32
forget .to(device) a641d5cdf82eb89c9034253785e80dd920c1f5d6 Thai Thien 2020-04-30 12:25:22
reduce batch size of shb keepfull 044f6785fedcf442def0fdb24109413a79274289 Thai Thien 2020-04-30 12:20:21
ccnn v7 with ssim loss 40079e5354eef95e0e89cf23a7f2025ee362e232 Thai Thien 2020-04-30 12:16:19
local 1d4e08754bd493e76cf09076de7c76adf1a30a0a Thai Thien 2020-04-30 12:10:15
mse-ssim loss function 052de7c879bff7690a7cfc1905c8376bf8605c45 Thai Thien 2020-04-30 11:41:24
delete misinformation note ca07533d585def04eec086685f2e72eacc330ddc Thai Thien 2020-04-29 16:59:47
typo 3505077301af5349665f12121862db2512ad450a Thai Thien 2020-04-29 16:59:03
change, adapt, survive 6755c8375302af2146cb63adc967631d53a7b1c8 Thai Thien 2020-04-29 16:48:10
Commit 3fea48c4642f62ab92accbe2493b6275d68c1b02 - no more batch timer
Author: Thai Thien
Author date (UTC): 2020-04-30 19:04
Committer name: Thai Thien
Committer date (UTC): 2020-04-30 19:04
Parent(s): 9e21a25fd4194e99c7899f7c6b6f04d51b432c61
Signer:
Signing key:
Signing status: N
Tree: 33c86d8113e537219f254b04298bc4b42ddad84d
File Lines added Lines deleted
train_compact_cnn.py 8 8
File train_compact_cnn.py changed (mode: 100644) (index 0c09543..a3f3a6e)
... ... if __name__ == "__main__":
111 111
112 112 # timer # timer
113 113 train_timer = Timer(average=True) # time to train whole epoch train_timer = Timer(average=True) # time to train whole epoch
114 batch_timer = Timer(average=True) # every batch
114 # batch_timer = Timer(average=True) # every batch
115 115 evaluate_timer = Timer(average=True) evaluate_timer = Timer(average=True)
116 116
117 batch_timer.attach(trainer,
118 start =Events.EPOCH_STARTED,
119 resume =Events.ITERATION_STARTED,
120 pause =Events.ITERATION_COMPLETED,
121 step =Events.ITERATION_COMPLETED)
117 # batch_timer.attach(trainer,
118 # start =Events.EPOCH_STARTED,
119 # resume =Events.ITERATION_STARTED,
120 # pause =Events.ITERATION_COMPLETED,
121 # step =Events.ITERATION_COMPLETED)
122 122
123 123 train_timer.attach(trainer, train_timer.attach(trainer,
124 124 start =Events.EPOCH_STARTED, start =Events.EPOCH_STARTED,
 
... ... if __name__ == "__main__":
159 159 experiment.log_metric("train_loss", metrics['loss']) experiment.log_metric("train_loss", metrics['loss'])
160 160 experiment.log_metric("lr", get_lr(optimizer)) experiment.log_metric("lr", get_lr(optimizer))
161 161
162 experiment.log_metric("batch_timer", batch_timer.value())
162 #experiment.log_metric("batch_timer", batch_timer.value())
163 163 experiment.log_metric("train_timer", train_timer.value()) experiment.log_metric("train_timer", train_timer.value())
164 164
165 print("batch_timer ", batch_timer.value())
165 #print("batch_timer ", batch_timer.value())
166 166 print("train_timer ", train_timer.value()) print("train_timer ", train_timer.value())
167 167
168 168 @trainer.on(Events.EPOCH_COMPLETED) @trainer.on(Events.EPOCH_COMPLETED)
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