v0.13.5

Detector fixes:

• Make sure the detector does not end up in an endless retry loop if initialization of the model fails

v0.13.4

Detector config:

• Set sio data limit to 20 MB

v0.13.3

#53 - Fixes a major bug in loop communication where the uploaded files in a batch upload all had
the same name. This caused the learning loop to import the same file up
to 20 times while losing up to 19 uploaded images.

v0.13.2

• fixes a bug where the creation date parameter was ignored on the socketio detect endpoint (#49)

v0.13.1

Detector fix:

Improve behavior when the target model is changed to speed up the change between already downloaded models. Downloaded models will only be compiled if necessary.

v0.13.0

Detector SIO Interface:

• Add endpoints to control versioning mode and get current versioning info (set_model_version_mode, get_model_version)
• Add endpoints to get about info (same response as the corresponding REST endpoint)
• Add endpoints to control outbox mode (get_outbox_mode, set_outbox_mode)
• Breaking change: All SIO responses are now dicts. This affects the response of the info endpoint which now returns {"status": "No model loaded"} if no model is loaded.

All Nodes:

• Changes of the node status are now logged with level INFO

Further Breaking Changes:

• Enums are moved to own module

v0.12.1

Improvements:

• Handle 429 (too many requests). When communicating with a learning loop instance via the REST interface the server may respond with 429 when it has to handle too many requests. We extended the loop communication module (used by all nodes) with a retry mechanism to retry the call several times with a delay between all attempts.

v0.12.0

Trainer Loop Communication Changes:

• No longer send redundant fields (like the hyperparameters, ... ) with trainer updates (they are send with the training updates)
• Change type of hyperparameters to a dict.
• Add fields trainer_name and best_epoch to DC TrainingOut - used to sync the training

Trainer - Internal Changes:

• Removes the internal DC TrainingData and moves its fields (image_data, categories, skipped_image_count, hyperparameter) to the Training dataclass

v0.11.1

Fixes:

• fixes a field name when using custom creation date

v0.11.0

Detector Feature:

• When running inference using a detector node, the user is now able to provide a custom creation date which will be used as creation date when the image is uploaded to the learning loop

Breaking Change for Multiple Node Types:

• The data class Detections is now split into:
  • Detections - used by the Trainer Node to upload detections for existing images after training is finished
  • ImageMetadata - used by the Detector when uploading a pair of image, image metadata and detections