waldo-seg / waldo

image-segmentation and text-localization
Apache License 2.0
13 stars 13 forks source link

WIP: config class draft #19

Closed YiwenShaoStephen closed 6 years ago

danpovey commented 6 years ago

ok, no problem.

On Wed, May 9, 2018 at 1:17 PM, Yiwen Shao notifications@github.com wrote:

@YiwenShaoStephen commented on this pull request.

In scripts/waldo/config.py https://github.com/waldo-seg/waldo/pull/19#discussion_r187112151:

  • self.config_dict.TRAIN.EPOCHS = 10
  • self.config_dict.TRAIN.START_EPOCH = 0
  • self.config_dict.TRAIN.BATCH_SIZE = 16
  • self.config_dict.TRAIN.INIT_LEARNING_RATE = 0.01
  • self.config_dict.TRAIN.MOMENTUM = 0.9
  • self.config_dict.TRAIN.NESTEROV = True
  • self.config_dict.TRAIN.WEIGHT_DECAY = 5e-4
  • network architecture

  • self.config_dict.NET = edict()
  • self.config_dict.NET.DEPTH = 5
  • self.config_dict.NET.NUM_OFFSETS = 15
  • self.config_dict.NET.NUM_CLASSES = 2
  • image

  • self.config_dict.IMAGE = edict()

Got it. BTW, I have a final project due tonight. Can I do this after the due if it's not urgent? Or you can ask someone else to do it if it's in an urgent need. Sorry for that!

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/waldo-seg/waldo/pull/19#discussion_r187112151, or mute the thread https://github.com/notifications/unsubscribe-auth/ADJVu0goGdV2cI4FxiGcVAeN_6pfIzKeks5twySSgaJpZM4T3rLu .