Need to specify TF1.x

[kaveh3000]

This will not run on TF2 due to the fact that tf.train.AdamOptimizer is not available in TF2.
Hence, I am proposing this minor change to ensure TF1.x is selected when the model is built.

[PapaMarc]

see #24 for v2 compat

[tjacobconley1]

I got an error about the modified code but still don't completely understand it [image: Igotanerrorabout the modified code but still dont completely understand it.JPG]

…

On Tue, Feb 23, 2021 at 10:29 PM PapaMarc ***@***.***> wrote: see #24 <#24> for v2 compat — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#19 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AHCJEBPDYM2WWWYWZPYPDZ3TARXALANCNFSM4QEE5OZA> .

-- *Tyler J. Conley* *[email protected] <[email protected]>*

[PapaMarc]

Then change simply facilitates running the existing code with the v2 tensorflow library v. the v1 library against which it was originally written. The v2 api changed the optimizer related function, so for those of us now running it for the first time with a v2 version of tensorflow, it does not run as (originally written)... but against the v2 version of the library, with the update, it performs as expected. From: tjacobconley1 [mailto:[email protected]] Sent: Wednesday, February 24, 2021 11:50 AM To: lexfridman/mit-deep-learning <[email protected]> Cc: PapaMarc <[email protected]>; Comment <[email protected]> Subject: Re: [lexfridman/mit-deep-learning] Need to specify TF1.x (#19) I got an error about the modified code but still don't completely understand it [image: Igotanerrorabout the modified code but still dont completely understand it.JPG]

…

On Tue, Feb 23, 2021 at 10:29 PM PapaMarc ***@***.***> wrote: see #24 <#24> for v2 compat - You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#19 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AHCJEBPDYM2WWWYWZPYPDZ3TARXALANCNFSM4QEE5OZA> .

-- *Tyler J. Conley* *[email protected] <[email protected]>* - You are receiving this because you commented. Reply to this email directly, view it on GitHub<https://apac01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Flexfridman%2Fmit-deep-learning%2Fpull%2F19%23issuecomment-784731637&data=04%7C01%7C%7C1c846e5ceff64bf7878608d8d8773bd9%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637497353937144204%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=XfPoAXcA%2B3MSAa0HdSXbmjJ2G%2BP2kkLD%2BRnyYDs3LKA%3D&reserved=0>, or unsubscribe<https://apac01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FAH3O7JLIAF5DGSNUCS742ODTARZNZANCNFSM4QEE5OZA&data=04%7C01%7C%7C1c846e5ceff64bf7878608d8d8773bd9%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637497353937154203%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=x5OYD%2F9LxRHo9kMsZP1jl5%2F%2FtZUtXRXgoD%2Fg%2FAbiAVk%3D&reserved=0>.