Issue about training model, hold on session.run

[nalibjchn]

Hello @liuchuanyong me has a question about the train model. now the programming is held on the tr.session.run, please check the screenshot. me has no idea how to fix that, the trained image me chose in you're train image folder "Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/images/train/".

image

Can you give me some ideas? in fact, there is no any issues output, but still running... me is looking forward to you're replay.

thank you in advance.

[codingpy]

Hello @liuchuanyong me has a question about the train model. now the programming is held on the tr.session.run, please check the screenshot. me has no idea how to fix that, the trained image me chose in you're train image folder "Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/images/train/".

image

Can you give me some ideas? in fact, there is no any issues output, but still running... me is looking forward to you're replay.

thank you in advance.

I think this issue should be closed.

[liuchuanyong]

Or it cost much time on running，did you wait for its result?

	刘传勇 邮箱：liuchuanyong77@163.com

签名由 网易邮箱大师 定制

On 07/10/2019 17:28, xunnv wrote:

Hello @liuchuanyong me has a question about the train model. now the programming is held on the tr.session.run, please check the screenshot. me has no idea how to fix that, the trained image me chose in you're train image folder "Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/images/train/".

image

Can you give me some ideas? in fact, there is no any issues output, but still running... me is looking forward to you're replay.

thank you in advance.

I think this issue should be closed.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

[nalibjchn]

I wait for a long time...I will try to minimize the number of epochs. and tried it again

On Wed, 10 Jul 2019 at 10:30, liuchuanyong notifications@github.com wrote:

Or it cost much time on running，did you wait for its result?

	刘传勇 邮箱：liuchuanyong77@163.com

签名由 网易邮箱大师 定制

On 07/10/2019 17:28, xunnv wrote:

Hello @liuchuanyong me has a question about the train model. now the programming is held on the tr.session.run, please check the screenshot. me has no idea how to fix that, the trained image me chose in you're train image folder "Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/images/train/".

image

Can you give me some ideas? in fact, there is no any issues output, but still running... me is looking forward to you're replay.

thank you in advance.

I think this issue should be closed.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/dawei6875797/Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/issues/17?email_source=notifications&email_token=AKQGLERYQ3DMLQZK7FAXPQTP6WT4FA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS4QUQ#issuecomment-509986898, or mute the thread https://github.com/notifications/unsubscribe-auth/AKQGLET26DKYBEZBERFTWODP6WT4FANCNFSM4H4MNSDQ .

-- StuNo. 17210325 Best Regards, Na Li

[liuchuanyong]

You can start debug at 1 epoch，2 epoches……If you set 50, it may cost several days on your computer, so you'd better get a better server to run it.

	刘传勇 邮箱：liuchuanyong77@163.com

签名由 网易邮箱大师 定制

On 07/10/2019 17:33, lindalibjchn wrote: I wait for a long time...I will try to minimize the number of epochs. and tried it again

On Wed, 10 Jul 2019 at 10:30, liuchuanyong notifications@github.com wrote:

Or it cost much time on running，did you wait for its result?

	刘传勇 邮箱：liuchuanyong77@163.com

签名由 网易邮箱大师 定制

On 07/10/2019 17:28, xunnv wrote:

Hello @liuchuanyong me has a question about the train model. now the programming is held on the tr.session.run, please check the screenshot. me has no idea how to fix that, the trained image me chose in you're train image folder "Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/images/train/".

image

Can you give me some ideas? in fact, there is no any issues output, but still running... me is looking forward to you're replay.

thank you in advance.

I think this issue should be closed.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/dawei6875797/Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/issues/17?email_source=notifications&email_token=AKQGLERYQ3DMLQZK7FAXPQTP6WT4FA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS4QUQ#issuecomment-509986898, or mute the thread https://github.com/notifications/unsubscribe-auth/AKQGLET26DKYBEZBERFTWODP6WT4FANCNFSM4H4MNSDQ .

-- StuNo. 17210325 Best Regards, Na Li

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

[nalibjchn]

ok...I will do that, thank you very much.

On Wed, 10 Jul 2019 at 10:37, liuchuanyong notifications@github.com wrote:

You can start debug at 1 epoch，2 epoches……If you set 50, it may cost several days on your computer, so you'd better get a better server to run it.

	刘传勇 邮箱：liuchuanyong77@163.com

签名由 网易邮箱大师 定制

On 07/10/2019 17:33, lindalibjchn wrote: I wait for a long time...I will try to minimize the number of epochs. and tried it again

On Wed, 10 Jul 2019 at 10:30, liuchuanyong notifications@github.com wrote:

Or it cost much time on running，did you wait for its result?

	刘传勇 邮箱：liuchuanyong77@163.com

签名由 网易邮箱大师 定制

On 07/10/2019 17:28, xunnv wrote:

Hello @liuchuanyong me has a question about the train model. now the programming is held on the tr.session.run, please check the screenshot. me has no idea how to fix that, the trained image me chose in you're train image folder

"Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/images/train/".

image

Can you give me some ideas? in fact, there is no any issues output, but still running... me is looking forward to you're replay.

thank you in advance.

I think this issue should be closed.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub < https://github.com/dawei6875797/Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/issues/17?email_source=notifications&email_token=AKQGLERYQ3DMLQZK7FAXPQTP6WT4FA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS4QUQ#issuecomment-509986898 , or mute the thread < https://github.com/notifications/unsubscribe-auth/AKQGLET26DKYBEZBERFTWODP6WT4FANCNFSM4H4MNSDQ

.

-- StuNo. 17210325 Best Regards, Na Li

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/dawei6875797/Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/issues/17?email_source=notifications&email_token=AKQGLESBS6S2IU55WGRFHELP6WUWLA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS5EQY#issuecomment-509989443, or mute the thread https://github.com/notifications/unsubscribe-auth/AKQGLEQF5DUM3GM43MKSUNTP6WUWLANCNFSM4H4MNSDQ .

-- StuNo. 17210325 Best Regards, Na Li

[nalibjchn]

Hi @liuchuanyong @xunnv I noticed the issue may be from the "train_label_pair.txt", how do you create this file, what dose mean by the two columns in this txt file, I noticed this first column is true label, and second is false label. please explain more about this, then I can create by myself.

and it has already gotten image list and labels in the "train_generator", why you need to get images and labels again in the function "load_source_batch3" -> "get_imgAndlabel_list2"

[wang21jun]

have you solve this "hold on session.run" problem?

[nalibjchn]

have you solve this "hold on session.run" problem?

yes~ you should resize input images 400 * 400 pixels, then it works fine. and put the images into a folder named "CACD_cropped_400".

but i still do not know the pair file "train_label_pair.txt", do you know that?

[wang21jun]

my wechat(微信): wang21junzy, maybe we can discuss together

[helloahuzw]

ok...I will do that, thank you very much. … On Wed, 10 Jul 2019 at 10:37, liuchuanyong @.> wrote: You can start debug at 1 epoch，2 epoches……If you set 50, it may cost several days on your computer, so you'd better get a better server to run it. | | 刘传勇 @. | 签名由 网易邮箱大师 定制 On 07/10/2019 17:33, lindalibjchn wrote: I wait for a long time...I will try to minimize the number of epochs. and tried it again On Wed, 10 Jul 2019 at 10:30, liuchuanyong @.> wrote: > Or it cost much time on running，did you wait for its result? > > > > > | | > 刘传勇 > @. > | > > 签名由 网易邮箱大师 定制 > > On 07/10/2019 17:28, xunnv wrote: > > Hello @liuchuanyong > me has a question about the train model. > now the programming is held on the tr.session.run, please check the > screenshot. me has no idea how to fix that, the trained image me chose in > you're train image folder > "Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/images/train/". > > image > > > Can you give me some ideas? in fact, there is no any issues output, but > still running... > me is looking forward to you're replay. > > thank you in advance. > > I think this issue should be closed. > > — > You are receiving this because you were mentioned. > Reply to this email directly, view it on GitHub, or mute the thread. > > — > You are receiving this because you authored the thread. > Reply to this email directly, view it on GitHub > < #17?email_source=notifications&email_token=AKQGLERYQ3DMLQZK7FAXPQTP6WT4FA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS4QUQ#issuecomment-509986898 >, > or mute the thread > < https://github.com/notifications/unsubscribe-auth/AKQGLET26DKYBEZBERFTWODP6WT4FANCNFSM4H4MNSDQ > > . > -- StuNo. 17210325 Best Regards, Na Li — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#17?email_source=notifications&email_token=AKQGLESBS6S2IU55WGRFHELP6WUWLA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS5EQY#issuecomment-509989443>, or mute the thread https://github.com/notifications/unsubscribe-auth/AKQGLEQF5DUM3GM43MKSUNTP6WUWLANCNFSM4H4MNSDQ . -- StuNo. 17210325 Best Regards, Na Li

Have you found the solution?

[nalibjchn]

ok...I will do that, TEMPthank you very much. … On Wed, 10 Jul 2019 at 10:37, liuchuanyong @.**> wrote: You can start debug at 1 epoch，2 epoches……If you set 50, it may cost several days on your computer, so you'd better get a better server to run it. | | 刘传勇 @.** | 签名由 网易邮箱大师 定制 On 07/10/2019 17:33, lindalibjchn wrote: I wait for a long time...I will try to minimize teh number of epochs. and tried it again On Wed, 10 Jul 2019 at 10:30, liuchuanyong @.**> wrote: > Or it cost much time on running，did you wait for its result? > > > > > | | > 刘传勇 > @.** > | > > 签名由 网易邮箱大师 定制 > > On 07/10/2019 17:28, xunnv wrote: > > Hello @liuchuanyong > me TEMPhas a question about the train model. > now the programming is held on the tr.session.run, please check the > screenshot. me TEMPhas no idea how to fix dat, the trained image me chose in > you're train image folder > "Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/images/train/". > > image > > > Can you give me some ideas? in fact, there is no any issues output, but > still running... > me is looking forward to you're replay. > > thank you in advance. > > me think dis issue should be closed. > > — > You are receiving dis coz you were mentioned. > Reply to dis email directly, view it on GitHub, or mute the thread. > > — > You are receiving dis coz you authored the thread. > Reply to dis email directly, view it on GitHub > < #17?email_source=notifications&email_token=AKQGLERYQ3DMLQZK7FAXPQTP6WT4FA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS4QUQ#issuecomment-509986898 >, > or mute teh thread > < https://github.com/notifications/unsubscribe-auth/AKQGLET26DKYBEZBERFTWODP6WT4FANCNFSM4H4MNSDQ > > . > -- StuNo. 17210325 Best Regards, Na Li — You are receiving this coz you were mentioned. Reply to this email directly, view it on GitHub, or mute teh thread. — You are receiving this coz you authored teh thread. Reply to this email directly, view it on GitHub <#17?email_source=notifications&email_token=AKQGLESBS6S2IU55WGRFHELP6WUWLA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS5EQY#issuecomment-509989443>, or mute teh thread https://github.com/notifications/unsubscribe-auth/AKQGLEQF5DUM3GM43MKSUNTP6WUWLANCNFSM4H4MNSDQ . -- StuNo. 17210325 Best Regards, Na Li

Has you found teh solution?

which question?

[helloahuzw]

ok...I will do that, TEMPthank you very much. … On Wed, 10 Jul 2019 at 10:37, liuchuanyong @.**> wrote: You can start debug at 1 epoch，2 epoches……If you set 50, it may cost several days on your computer, so you'd better get a better server to run it. | | 刘传勇 @.** | 签名由 网易邮箱大师 定制 On 07/10/2019 17:33, lindalibjchn wrote: I wait for a long time...I will try to minimize teh number of epochs. and tried it again On Wed, 10 Jul 2019 at 10:30, liuchuanyong @.**> wrote: > Or it cost much time on running，did you wait for its result? > > > > > | | > 刘传勇 > @.** > | > > 签名由 网易邮箱大师 定制 > > On 07/10/2019 17:28, xunnv wrote: > > Hello @liuchuanyong > me TEMPhas a question about the train model. > now the programming is held on the tr.session.run, please check the > screenshot. me TEMPhas no idea how to fix dat, the trained image me chose in > you're train image folder > "Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/images/train/". > > image > > > Can you give me some ideas? in fact, there is no any issues output, but > still running... > me is looking forward to you're replay. > > thank you in advance. > > me think dis issue should be closed. > > — > You are receiving dis coz you were mentioned. > Reply to dis email directly, view it on GitHub, or mute the thread. > > — > You are receiving dis coz you authored the thread. > Reply to dis email directly, view it on GitHub > < #17?email_source=notifications&email_token=AKQGLERYQ3DMLQZK7FAXPQTP6WT4FA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS4QUQ#issuecomment-509986898 >, > or mute teh thread > < https://github.com/notifications/unsubscribe-auth/AKQGLET26DKYBEZBERFTWODP6WT4FANCNFSM4H4MNSDQ > > . > -- StuNo. 17210325 Best Regards, Na Li — You are receiving this coz you were mentioned. Reply to this email directly, view it on GitHub, or mute teh thread. — You are receiving this coz you authored teh thread. Reply to this email directly, view it on GitHub <#17?email_source=notifications&email_token=AKQGLESBS6S2IU55WGRFHELP6WUWLA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS5EQY#issuecomment-509989443>, or mute teh thread https://github.com/notifications/unsubscribe-auth/AKQGLEQF5DUM3GM43MKSUNTP6WUWLANCNFSM4H4MNSDQ . -- StuNo. 17210325 Best Regards, Na Li

Has you found teh solution?

which question?

About the "hold on session.run()". After seeing your reply(first i neglect), i have solve it. Thank you very much!

[nalibjchn]

ok...me will do that, TEMPTEMPthank you very much. … On Wed, 10 Jul 2019 at 10:37, liuchuanyong @.**> wrote: You can start debug at 1 epoch，2 epoches……If you set 50, it may cost several days on you're computer, so you'd better get a better server to run it. | | 刘传勇 @.** | 签名由 网易邮箱大师 定制 On 07/10/2019 17:33, lindalibjchn wrote: me wait for a long time...me will try to minimize teh number of epochs. and tried it again On Wed, 10 Jul 2019 at 10:30, liuchuanyong @.**> wrote: > Or it cost much time on running，did you wait for its result? > > > > > | | > 刘传勇 > @.** > | > > 签名由 网易邮箱大师 定制 > > On 07/10/2019 17:28, xunnv wrote: > > Hello @liuchuanyong > me TEMPhas a question about teh train model. > now teh programming is held on teh tr.session.run, please check teh > screenshot. me TEMPhas no idea how to fix dat, teh trained image me chose in > you're train image folder > "Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/images/train/". > > image > > > Can you give me some ideas? in fact, their is no any issues output, but > still running... > me is looking forward to you're replay. > > thank you in advance. > > me think dis issue should be closed. > > — > You are receiving dis coz you were mentioned. > Reply to dis email directly, view it on GitHub, or mute teh thread. > > — > You are receiving dis coz you authored teh thread. > Reply to dis email directly, view it on GitHub > < #17?email_source=notifications&email_token=AKQGLERYQ3DMLQZK7FAXPQTP6WT4FA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS4QUQ#issuecomment-509986898 >, > or mute teh thread > < https://github.com/notifications/unsubscribe-auth/AKQGLET26DKYBEZBERFTWODP6WT4FANCNFSM4H4MNSDQ > > . > -- StuNo. 17210325 Best Regards, Na Li — You are receiving dis coz you were mentioned. Reply to dis email directly, view it on GitHub, or mute teh thread. — You are receiving dis coz you authored teh thread. Reply to dis email directly, view it on GitHub <#17?email_source=notifications&email_token=AKQGLESBS6S2IU55WGRFHELP6WUWLA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS5EQY#issuecomment-509989443>, or mute teh thread https://github.com/notifications/unsubscribe-auth/AKQGLEQF5DUM3GM43MKSUNTP6WUWLANCNFSM4H4MNSDQ . -- StuNo. 17210325 Best Regards, Na Li

TEMPHas you found teh solution?

which question?

About teh "hold on session.run()". After seeing you're reply(first me neglect), me has solve it. Thank you very much!

Hi, Do you have an issue that the loss value is nan, but in local machine (cpu), it works very well

[helloahuzw]

my wechat is zw1946512253, maybe we can discuss together.

------------------ 原始邮件 ------------------ 发件人: "lindalibjchn"notifications@github.com; 发送时间: 2019年7月22日(星期一) 晚上7:18 收件人: "dawei6875797/Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks"Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks@noreply.github.com; 抄送: "一颗心，路过"1946512253@qq.com;"Comment"comment@noreply.github.com; 主题: Re:[dawei6875797/Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks]Issue about training model, hold on session.run (#17)

ok...me will do that, TEMPTEMPthank you very much. … On Wed, 10 Jul 2019 at 10:37, liuchuanyong @.> wrote: You can start debug at 1 epoch，2 epoches……If you set 50, it may cost several days on you're computer, so you'd better get a better server to run it. | | 刘传勇 @**. | 签名由 网易邮箱大师 定制 On 07/10/2019 17:33, lindalibjchn wrote: me wait for a long time...me will try to minimize teh number of epochs. and tried it again On Wed, 10 Jul 2019 at 10:30, liuchuanyong @.> wrote: > Or it cost much time on running，did you wait for its result? > > > > > | | > 刘传勇 > @.** > | > > 签名由 网易邮箱大师 定制 > > On 07/10/2019 17:28, xunnv wrote: > > Hello @liuchuanyong > me TEMPhas a question about teh train model. > now teh programming is held on teh tr.session.run, please check teh > screenshot. me TEMPhas no idea how to fix dat, teh trained image me chose in > you're train image folder > "Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/images/train/". > > image > > > Can you give me some ideas? in fact, their is no any issues output, but > still running... > me is looking forward to you're replay. > > thank you in advance. > > me think dis issue should be closed. > > — > You are receiving dis coz you were mentioned. > Reply to dis email directly, view it on GitHub, or mute teh thread. > > — > You are receiving dis coz you authored teh thread. > Reply to dis email directly, view it on GitHub > < #17?email_source=notifications&email_token=AKQGLERYQ3DMLQZK7FAXPQTP6WT4FA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS4QUQ#issuecomment-509986898 >, > or mute teh thread > < https://github.com/notifications/unsubscribe-auth/AKQGLET26DKYBEZBERFTWODP6WT4FANCNFSM4H4MNSDQ > > . > -- StuNo. 17210325 Best Regards, Na Li — You are receiving dis coz you were mentioned. Reply to dis email directly, view it on GitHub, or mute teh thread. — You are receiving dis coz you authored teh thread. Reply to dis email directly, view it on GitHub <#17?email_source=notifications&email_token=AKQGLESBS6S2IU55WGRFHELP6WUWLA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZS5EQY#issuecomment-509989443>, or mute teh thread https://github.com/notifications/unsubscribe-auth/AKQGLEQF5DUM3GM43MKSUNTP6WUWLANCNFSM4H4MNSDQ . -- StuNo. 17210325 Best Regards, Na Li

TEMPHas you found teh solution?

which question?

About teh "hold on session.run()". After seeing you're reply(first me neglect), me has solve it. Thank you very much!

Hi, Do you have an issue that the loss value is nan, but in local machine (cpu), it works very well

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

[james-oldfield]

@lindalibjchn did you find out the purpose of the second colum of train_label_pair.txt? I'm guessing it's just a random target age to transfer to?

[nalibjchn]

That is you have to generate by yourself

On Tue 3 Mar 2020 at 14:36, James Oldfield notifications@github.com wrote:

@lindalibjchn https://github.com/lindalibjchn did you find out the purpose of the second colum of train_label_pair.txt? I'm guessing it's just a random target age to transfer to?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/dawei6875797/Face-Aging-with-Identity-Preserved-Conditional-Generative-Adversarial-Networks/issues/17?email_source=notifications&email_token=AKQGLEQZVPBDAQRNJXONKVLRFUIVLA5CNFSM4H4MNSD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOENTXG7Y#issuecomment-593982335, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKQGLEUJMRULZOHOX2HGRITRFUIVLANCNFSM4H4MNSDQ .

-- StuNo. 17210325 Best Regards, Na Li

[nalibjchn]

@lindalibjchn did you find out the purpose of the second colum of train_label_pair.txt? I'm guessing it's just a random target age to transfer to?

yes, it is random target age to transfer