pipeline:window:cryolo:changelog

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
pipeline:window:cryolo:changelog [2019/09/30 16:07]
twagner
pipeline:window:cryolo:changelog [2020/05/25 10:17] (current)
twagner
Line 1: Line 1:
 +<note important>​
 +
 +**DOCUMENTATION OUTDATED**
 +
 +Find new versions here: https://​cryolo.readthedocs.io/​en/​latest/​
 +
 +</​note>​
 +
 +**crYOLO 1.6.1:**
 +  * Fixed a bug that was introduced with 1.5.5: Scaling of the anchor boxes was wrong. This leads to longer and unstable training and heavily affects the fine-tune mode. (Thanks to Jorge Jimenez de la Morena and Pablo Conesa)
 +  * Fixed a bug that leads to an exception (//​_tkinter.TclError:​ couldn'​t connect to display//) at the end of the training on cluster machines. (Thanks to Wolfgang Lugmayr)
 +
 +**crYOLO 1.6.0:**
 +  * In case of the general model, you can specify with %%--%%minsize MIN %%--%%maxsize MAX a minimum and maximum size. This will filter the particles according to their estimated size.
 +  * The estimated size and confidence distribution are now written in a new subfolder DISTR in your output folder. It will also write .csv files with a summary of the distributions.
 +  * In case of the general model, you don't need to specify the anchor size anymore.
 +  * With every run, crYOLO now writes the command used into the central log directory.
 +  * All log files (runfiles, commands, tensorflow) are now saved in the central log directory.
 +  * During training, the intermediate models now get a suffix "​_tmp"​. After training is finished they are renamed to the specified name in the configuration file (field: "​save_weights_name"​).
 +  * The boxmanager can now be started through the crYOLO GUI.
 +  * Fixed issue that the filament mode does not work with micrographs that were motion-corrected by unblur.
 +  * Fixed issue that the flaq %%--%%write_empty did not work for the filament mode.
 +  * Fixed issue that the minimum distance filter was not applied on particles in .cbox files.
 +  * Fixed issue with the evaluation tool that crashed if no particle can be found for a specific threshold.
 +
 +<hidden **Old crYOLO change logs**>
 +
 +**crYOLO 1.5.6:**
 +  * Fix installation issues
 +
 +**crYOLO 1.5.5:**
 +  * Added a the option %%--%%use_multithreading for training. If python multiprocessing leads to problems during training (e.g. freezing, dying workers) use multithreading instead of multiprocessing.
 +  * Boxmanager 1.2.8 integrates a low pass filter to make it training data creation easier.
 +  * Internal refactoring needed for boxmanger 1.2.8 changes.
 +
 +**crYOLO 1.5.4:**
 +  * Fixed systematic offset when picking K3 images using the general model.
 +  * Fixed crash when minimum distance filter is used (Thanks to Peter Van Blerkom)
 +  * Add "Print cmd" button to GUI. The command can then be copied to a submission script (Thanks to Wolfgang Lugmayr).
 +  * Fixed problem that crYOLO does not work if no filtering is used.
 +
 **crYOLO 1.5.3:** **crYOLO 1.5.3:**
-  * Fix another problem when downloading pretrained weights +  * Fixed another problem when downloading pretrained weights 
-  * Fix problem when opening the GUI on a computer without a GPU it crashes. This is especially problematic for clusters with GPU nodes.+  * Fixed problem when opening the GUI on a computer without a GPU it crashes. This is especially problematic for clusters with GPU nodes.
  
 **crYOLO 1.5.1:** **crYOLO 1.5.1:**
Line 12: Line 53:
   * Added a GUI for crYOLO (start with cryolo_gui.py)   * Added a GUI for crYOLO (start with cryolo_gui.py)
   * Optimized fast low-pass filtering pipeline: In consequence this, speeds up on-the-fly filtering by 175%, training by 75% and picking by 60%   * Optimized fast low-pass filtering pipeline: In consequence this, speeds up on-the-fly filtering by 175%, training by 75% and picking by 60%
-  * New monitor mode for prediction: When this option is activated, crYOLO will monitor your input folder. This especially useful for automation purposes. You can stop the monitor mode by writing an empty file with the name "​stop.cryolo"​ in the input directory. Just add --monitor in the command line or check the monitor box in GUI. +  * New monitor mode for prediction: When this option is activated, crYOLO will monitor your input folder. This especially useful for automation purposes. You can stop the monitor mode by writing an empty file with the name "​stop.cryolo"​ in the input directory. Just add %%--%%monitor in the command line or check the monitor box in GUI. 
   * Training is now merged into **one command** instead of two   * Training is now merged into **one command** instead of two
   * Add rotation as additional data augmentation   * Add rotation as additional data augmentation
Line 20: Line 61:
   * crYOLO will not allocate the complete GPU memory anymore.   * crYOLO will not allocate the complete GPU memory anymore.
   * Remove warmup as config file option. Please specify it with -w.   * Remove warmup as config file option. Please specify it with -w.
- 
- 
- 
-<hidden **Old crYOLO change logs**> 
  
 **crYOLO 1.4.1:** **crYOLO 1.4.1:**
  • pipeline/window/cryolo/changelog.1569852465.txt.gz
  • Last modified: 2019/09/30 16:07
  • by twagner