User Tools

Site Tools


pipeline:window:cryolo:training

Differences

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

Link to this comparison view

Next revision
Previous revision
pipeline:window:cryolo:training [2019/09/17 11:10]
twagner created
pipeline:window:cryolo:training [2020/03/16 15:21] (current)
twagner
Line 3: Line 3:
 nvidia-smi nvidia-smi
 </code> </code>
-For this tutorial, we assume that you have either a single GPU or want to use GPU 0. In the GUI you have to fille the mandatory fields: +For this tutorial, we assume that you have either a single GPU or want to use GPU 0. 
-{{ :pipeline:window:cryolo:cryolo_training.png?700 |}} +
- +
-The default number of warmup epochs is fine as long you don't want to refine an existing model. During the warmup training epochs it will not try to estimate the size of your particle, which helps crYOLO to converge.+
  
 <note tip> <note tip>
-Before you start the training, you might want to change the GPU. By default, crYOLO will use GPU 0. The following command will show the status of all GPUs+**Use a different or multiple GPUs**
-<code> +
-nvidia-smi +
-</code>+
  
-In //"Optional arguments"// tab you can change the GPU that should be used by crYOLO. If you have multiple (e.g GPU 0 and GPU 1) you could also use both by set the GPU argument to '0 1'+In the //"Optional arguments"// tab you can change the GPU that should be used by crYOLO. If you have multiple GPUs (e.g. nvidia-smi lists GPU 0 and GPU 1) you can also use both by setting the GPU argument to '0 1'
 </note> </note>
 +
 +
 +In the GUI you have to fill in the mandatory fields:
 +{{ :pipeline:window:cryolo:cryolo_training_202003.png?700 |}}
 +
 +The default number of warmup epochs((One epoch is a complete pass through the training data.)) is fine as long as you don't want to refine an existing model. During the warmup training epochs it will not try to estimate the size of your particle, which helps crYOLO to converge.
  
 <note tip> <note tip>
-When you start the training, it will stop when the "loss" metric on the validation data does not improve 10 times in a row. This is typically enough. In case want to give the training more time to find the best model you might want to increase the "not changed in a row" parameter to, for example, 15 by setting the //early// argument in the //"Optional arguments"// to 15.+**When does crYOLO stop the training?** 
 + 
 +When you start the training, it will stop when the "loss" metric on the validation data does not improve 10 times in a row. This is typically enough. In case you want to give the training more time to find the best model can increase the "not changed in a row" parameter to a higher value by setting the //early// argument in the //"Optional arguments"// to, for example, 15.
 </note> </note>
  
-Now press the Start button to start the training. The final model will be written to disk as specified in //saved_weights_name// in your configuration file.+The final model will be written to disk as specified in //saved_weights_name// in your configuration file.
  
-<hidden **Train crYOLO using the command line**>+<html> 
 +<div style="background-color: #cfc ; padding: 10px; border: 1px solid green;">  
 +<b>&#9658; Now press the [Start] button to start the training. </b> 
 +</div> 
 +</html> 
 + 
 +<html> 
 +<div style="background-color: LightCyan ; padding: 10px; border: 1px solid Black;">  
 +<b>Alternative: Train crYOLO using the command line</b> 
 +</div> 
 +</html> 
 + 
 + 
 +<hidden initialState="visible">
 Navigate to the folder with ''config_cryolo.json'' file, ''train_image'' folder, etc. Navigate to the folder with ''config_cryolo.json'' file, ''train_image'' folder, etc.
  
Line 29: Line 44:
  
 <code> <code>
-cryolo_train.py -c config.json -w 5 -g 0+cryolo_train.py -c config_cryolo.json -w 5 -g 0
 </code> </code>
  
 The final model file will be written to disk. The final model file will be written to disk.
  
 +</hidden>
  
- +<html
-<code+<div style="background-color: LightCyan ; padding: 10px; border: 1px solid Black;">  
-cryolo_train.py -c config.json -w 3 -g 0 -e 15 +<b> </b
-</code+</div> 
- +</html>
-to the training command. +
-</hidden>+
pipeline/window/cryolo/training.1568711426.txt.gz · Last modified: 2019/09/17 11:10 by twagner