szha commented on a change in pull request #9633: Gluon image-classification
example improvement
URL: https://github.com/apache/incubator-mxnet/pull/9633#discussion_r164932256
##########
File path: example/gluon/image_classification.py
##########
@@ -62,63 +75,85 @@
help='enable batch normalization or not in vgg. default is
false.')
parser.add_argument('--use-pretrained', action='store_true',
help='enable using pretrained model from gluon.')
+parser.add_argument('--prefix', default='', type=str,
+ help='path to checkpoint prefix, default is current
working dir')
+parser.add_argument('--start-epoch', default=0, type=int,
+ help='starting epoch, 0 for fresh training, > 0 to resume')
+parser.add_argument('--resume', type=str, default='',
+ help='path to saved weight where you want resume')
+parser.add_argument('--lr-factor', default=0.1, type=float,
+ help='learning rate decay ratio')
+parser.add_argument('--lr-steps', default='30,60,90', type=str,
+ help='list of learning rate decay epochs as in str')
+parser.add_argument('--dtype', default='float32', type=str,
+ help='data type, float32 or float16 if applicable')
+parser.add_argument('--save-frequency', default=10, type=int,
+ help='model save frequent, best model will always be
saved')
parser.add_argument('--kvstore', type=str, default='device',
help='kvstore to use for trainer/module.')
-parser.add_argument('--log-interval', type=int, default=50, help='Number of
batches to wait before logging.')
+parser.add_argument('--log-interval', type=int, default=50,
+ help='Number of batches to wait before logging.')
parser.add_argument('--profile', action='store_true',
help='Option to turn on memory profiling for front-end, '\
'and prints out the memory usage by python function
at the end.')
+parser.add_argument('--top-k', type=int, default=0, help='add top-k metric if
> 1')
Review comment:
there are many options already, which could affect its usability. should we
just hard-code this?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on 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
