Finetuner

Added

- Support pair-wise score document construction from CSV. ([696](https://github.com/jina-ai/finetuner/pull/696))

Removed

Changed

- Refactor `load_finetuning_dataset` into CSV handlers. ([696](https://github.com/jina-ai/finetuner/pull/696))

- Unify all models names into `name-size-lang` format. ([700](https://github.com/jina-ai/finetuner/pull/700))

- Do not download pre-trained weights when user downloads the artifact. ([706](https://github.com/jina-ai/finetuner/pull/706))

Fixed

Docs

- Add query-document score into CSV data preparation. ([697](https://github.com/jina-ai/finetuner/pull/697))

- Add `CosineSimilarityLoss` into advanced loss section. ([697](https://github.com/jina-ai/finetuner/pull/697))

- Add layer-wise learning rate decay into advanced configurations. ([697](https://github.com/jina-ai/finetuner/pull/697))

Added

- Add support for batch size scaling. ([691](https://github.com/jina-ai/finetuner/pull/691))

- Add functions to retrieve evaluation metrics and example results. ([687](https://github.com/jina-ai/finetuner/pull/687))

Removed

Changed

- Remove unit test and integration test from CD. ([686](https://github.com/jina-ai/finetuner/pull/686))

Fixed

Docs

Added

- Add support for learning rate schedulers. ([679](https://github.com/jina-ai/finetuner/pull/679))

Removed

Changed

- Remove duplicated documents when parsing unlabeled CSV files. ([678](https://github.com/jina-ai/finetuner/pull/678))

- The `scheduler_step` options is now part of `scheduler_options`. ([679](https://github.com/jina-ai/finetuner/pull/679))

Fixed

Docs

- Update documentation on creating training data. ([678](https://github.com/jina-ai/finetuner/pull/678))

- Add notebook to demonstrate use of `ArcFaceLoss`. ([680](https://github.com/jina-ai/finetuner/pull/680))

- Add section on GeM pooling to advanced topics. ([684](https://github.com/jina-ai/finetuner/pull/684))

Added

- Add support for new loss and pooling options to the `finetuner.fit` method. ([664](https://github.com/jina-ai/finetuner/pull/664))

- Add folder for example CSV files. ([663](https://github.com/jina-ai/finetuner/pull/663))

- Add communication between remote-ci job and the pr that triggered it. ([642](https://github.com/jina-ai/finetuner/pull/642))

- Support continuing training from an artifact of a previous run. ([668](https://github.com/jina-ai/finetuner/pull/668))

Removed

Changed

- Use github token provided by dispatcher when running remote-ci. ([640](https://github.com/jina-ai/finetuner/pull/640))

Fixed

- Use python 3.8 in Github actions. ([659](https://github.com/jina-ai/finetuner/pull/659))

- Add proper CSV file for image-image case. ([667](https://github.com/jina-ai/finetuner/pull/667))

- Fix problems with login function in notebooks by bumping hubble version. ([672](https://github.com/jina-ai/finetuner/pull/672))

- Fix URL construction. ([672](https://github.com/jina-ai/finetuner/pull/672))

Docs

- Add page on loss and pooling to `advanced-topics`. ([664](https://github.com/jina-ai/finetuner/pull/664))

- Remove ResNet backbone support for clip fine-tuning. ([662](https://github.com/jina-ai/finetuner/pull/662))

- Add efficientnet b7 as a new image to image search backbone. ([662](https://github.com/jina-ai/finetuner/pull/662))

- Fix typos, duplicate paragraphs, and wrong formulations. ([666](https://github.com/jina-ai/finetuner/pull/666))

- Add list of articles to README and docs. ([669](https://github.com/jina-ai/finetuner/pull/669))

- Removed section on GeM pooling from advanced topics. ([676](https://github.com/jina-ai/finetuner/pull/676))

Added

- Add `val_split` parameter to `fit` function. ([624](https://github.com/jina-ai/finetuner/pull/624))

- Add `core-ci` workflow to remotely run the ci of finetuner-core. ([628](https://github.com/jina-ai/finetuner/pull/628))

- Add support for 3d meshes to `build_finetuning_dataset`. ([638](https://github.com/jina-ai/finetuner/pull/638))

Removed

- Remove `cpu` parameter from `create_run` function. ([631](https://github.com/jina-ai/finetuner/pull/631))

- Remove `notebook_login` function. ([631](https://github.com/jina-ai/finetuner/pull/631))

- Remove support for python 3.7 ([653](https://github.com/jina-ai/finetuner/pull/653))

Changed

- Adjust Finetuner based on API changes for Jina AI Cloud. ([637](https://github.com/jina-ai/finetuner/pull/637))

- Change default `experiment_name` from current working dir to `default`. ([637](https://github.com/jina-ai/finetuner/pull/637))

- Use github token provided by dispatcher when running remote-ci. ([640](https://github.com/jina-ai/finetuner/pull/640))

Fixed

- Correctly infer the type of models created using `get_model` in the `build_encoding_dataset` function. ([623](https://github.com/jina-ai/finetuner/pull/623))

Docs

- Add before and after section to the example notebooks. ([622](https://github.com/jina-ai/finetuner/pull/622))

- Align text-to-image notebook with its corresponding markdown file. ([621](https://github.com/jina-ai/finetuner/pull/621))

- Change hint in notebooks to use `load_uri_to_blob` instead of `load_uri_to_image_tensor`. ([625](https://github.com/jina-ai/finetuner/pull/625))

- Copyedit `README.md`, changes to language but not contents. ([635](https://github.com/jina-ai/finetuner/pull/635))

- Add multilingual clip colab to readme. ([620](https://github.com/jina-ai/finetuner/pull/620))

- Add tutorial for mesh-to-mesh search. ([638](https://github.com/jina-ai/finetuner/pull/638))

- Add documentation for PointNet++ model and handling 3D mesh dataset. ([638](https://github.com/jina-ai/finetuner/pull/638))

- Add `finetuner` namespace to artifact names in the documentation. ([649](https://github.com/jina-ai/finetuner/pull/649))

- Rewrite M-CLIP notebook to use German fashion dataset. ([643](https://github.com/jina-ai/finetuner/pull/643))

- New advanced topics section. ([643](https://github.com/jina-ai/finetuner/pull/643))

- Improve developer reference. ([643](https://github.com/jina-ai/finetuner/pull/643))

- Improve walkthrough sections. ([643](https://github.com/jina-ai/finetuner/pull/643))

- Add hints to escape common to prepare csv training data. ([655](https://github.com/jina-ai/finetuner/pull/655))

Added

- Allow user to control `num_items_per_class` to sample to each batch. ([614](https://github.com/jina-ai/finetuner/pull/614))

Removed

Changed

- Update commons and stubs versions. ([618](https://github.com/jina-ai/finetuner/pull/618))

Fixed

- Valid configuration of `num_items_per_class`. ([618](https://github.com/jina-ai/finetuner/pull/618))

Docs

- Add notebook for multilingual CLIP models. ([611](https://github.com/jina-ai/finetuner/pull/611))

- Improve `describe_models` with `task` to better organize list of backbones. ([610](https://github.com/jina-ai/finetuner/pull/610))

- Add documentation on using the evaluation callback for CLIP (multiple models). ([615](https://github.com/jina-ai/finetuner/pull/615))

- Ignore `callback` module in apidoc. ([614](https://github.com/jina-ai/finetuner/pull/614))