MalteEbner
MalteEbner
Review this together with the changed API specs: https://github.com/lightly-ai/lightly-core/pull/1834 ## Changes - generates code from the new API specs in the linked PR - includes a unittest checking that the...
generation branch: https://github.com/lightly-ai/lightly-core/compare/malte-lig-455-use-openapi-generator-instead-of-swagger-api-part ## Changes - generates the new specs from openAPI - adapts the corresponding imports, e.g. to`from lightly.openapi_generated.swagger_client.model.docker_run_data import DockerRunData` ## Partly changed - Partly: changes the attributes...
## Description There is currently no documentation on creating custom metadata in the coco format easily. Creating it by hand is not easy, as you need to take care of...
Currently we have many different user groups for the lightly platform. Of course, some users may be part of multiple of these groups. A. potential customers, who want to evaluate...
## Current state The unmocked tests for CLI commands also used in the CI cover already quite some cases. https://github.com/lightly-ai/lightly/blob/master/tests/UNMOCKED_end2end_tests/run_all_unmocked_tests.sh We also have CI script for the CLI commands magic...
## Description To be able to fully script lightly workflows and automate them, we need to be able to get the output of lightly cli commands. This can be done...
## Current state: Currently, we offer 3 levels of SSL parts: - low-level: the separate building blocks such as a projection head, nn memory bank, backbone creation https://github.com/lightly-ai/lightly/blob/8dde4d68c0598758bfd05c05f225bf754d9f1ed5/lightly/models/modules/heads.py#L12 - mid-level:...
## Description: The tutorial tries to evaluate whether the embeddings are good by looking at it's neighbours: It computes for each example image the class distribution of neighbour images, which...
Currently the unittests create fake datasets and embeddings at several different locations and in serveral different ways. This could be reworked to have only one function for each. Locations: -...
We should follow one single approach how to set variables like the `token` or `dataset_id` in the active learning tutorials (and possibly also other tutorials): ## Approach A: https://github.com/lightly-ai/lightly/blob/c437183b8726b20d4bb5a7c97e02d3fdf21b14c6/docs/source/tutorials_source/platform/tutorial_active_learning_detectron2.py#L130-L140 The...