KhurramPirov commented on issue #15496: Loaded pretrained model but train 
accuracy starts from zero,
URL: 
https://github.com/apache/incubator-mxnet/issues/15496#issuecomment-511388284
 
 
   And pyfile
   
   пн, 15 июл. 2019 г. в 15:21, Пиров Хуррам Хуршедович <
   [email protected]>:
   
   > Hi! I attached an example for mnist_code, just train for some checkpoint
   > and then try to load it. All in the  notebook, I can't get the same
   > accuracy after loading from checkpoints. Please check my notebook. The same
   > bug is in my main code.
   >
   > пт, 12 июл. 2019 г. в 14:08, Rishik Mourya <[email protected]>:
   >
   >> Below code should work
   >>
   >> model.save_checkpoint('model', epoch = 0)
   >>
   >> sym, arg_params, aux_params = mx.model.load_checkpoint('model', epoch = 0)
   >>
   >> model = mx.mod.Module(sym, context = mx.gpu())
   >> model.bind(data_shapes = train_data.provide_data, label_shapes = 
train_data.provide_label)
   >> model.set_params(arg_params, aux_params)
   >>
   >> —
   >> You are receiving this because you were mentioned.
   >> Reply to this email directly, view it on GitHub
   >> 
<https://github.com/apache/incubator-mxnet/issues/15496?email_source=notifications&email_token=AH2IALWXSFZS2HGDQC5XVTLP7BQ3HA5CNFSM4H7EN2K2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZZOQQQ#issuecomment-510847042>,
   >> or mute the thread
   >> 
<https://github.com/notifications/unsubscribe-auth/AH2IALUDS2I4EIDGSB4CI5TP7BQ3HANCNFSM4H7EN2KQ>
   >> .
   >>
   >
   >
   > --
   > Khurram Pirov
   > skype: matrixrextended
   > mob: +7-916-439-53-76
   >
   >
   
   -- 
   Khurram Pirov
   skype: matrixrextended
   mob: +7-916-439-53-76
   

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
[email protected]


With regards,
Apache Git Services

Reply via email to