User Tools

Site Tools


cryolo_config

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
Next revision Both sides next revision
cryolo_config [2018/12/20 10:35]
twagner
cryolo_config [2018/12/20 18:00]
twagner
Line 2: Line 2:
 {{ :config_explanation.png?500|}} {{ :config_explanation.png?500|}}
 The config file is organized in the sections model, training and validation. The config file is organized in the sections model, training and validation.
-In the following your find a description of each entry.+In the following you find a description of each entry.
  
 **Model section:** **Model section:**
Line 30: Line 30:
   * #12 ''batch_size'': Specified the number of images crYOLO process in parallel during training. Strongly depending on the memory of your graphic card. 6 should be fine for GPUs with 8GB memory. You can increase in case you have more memory or decrease if you have memory problems. Bigger batches tend to improve convergence and even the final error.   * #12 ''batch_size'': Specified the number of images crYOLO process in parallel during training. Strongly depending on the memory of your graphic card. 6 should be fine for GPUs with 8GB memory. You can increase in case you have more memory or decrease if you have memory problems. Bigger batches tend to improve convergence and even the final error.
  
-  * #13 ''learning_rate'':+  * #13 ''learning_rate'': Defines the step size during training. Default should be kept.
  
-  * #14 ''nb_epoch'':+  * #14 ''nb_epoch'': Maximum number of epochs the network will train. I basically never reach this number, as crYOLO stops training if it recognize that the validation loss is not improving anymore.
  
-  * #15 ''object_scale'':+  * #15 ''object_scale'': Penality scaling factor for missing picking particles.
  
-  * #16 ''no_object_scale'':+  * #16 ''no_object_scale'': Penality scaling factor for picking background.
  
-  * #17 ''coord_scale'':+  * #17 ''coord_scale'': Penality scaling factor for errors in estimating the correct position.
  
-  * #18 ''class_scale'':+  * #18 ''class_scale'': Irrelevant, as crYOLO only has the "class" "particle".
  
-  * #19 ''log_path'':+  * #19 ''log_path'': Path to folder. During training, crYOLo saves there some logs for visualization in tensorboard. Tensorboard is used to visualize curves for training and validation loss.
  
-  * #20 ''saved_weights_name'':+  * #20 ''saved_weights_name'': Everytime the network improves in terms of validation loss, it will save the model into the file specified here.
  
-  * #21 ''debug'':+  * #21 ''debug'': If true, the network will provide several statistics during training.
  
 **Validation section:** **Validation section:**
  
-  * #22 ''valid_image_folder'':+  * #22 ''valid_image_folder'': If not specified, crYOLO will simply select 20% of the training data for validation. However it is possible to specify to use specific images for validation. This should be the path to folder containing these files.
  
-  * #23 ''valid_annot_folder'':+  * #23 ''valid_annot_folder'': If not specified, crYOLO will simply select 20% of the training data for validation. However it is possible to specify to use specific images for validation. This should be the path to folder containing these validation box files.
  
-  * #24 ''valid_times'':+  * #24 ''valid_times'': How often each image is presented the network during validation. 1 should be kept.
cryolo_config.txt ยท Last modified: 2019/10/11 17:51 by twagner