Finetuner Update 0.7.2

Finetuner makes neural network fine-tuning easier and faster by streamlining the workflow and handling all the complexity and infrastructure requirements in the cloud. With Finetuner, one can easily enhance the performance of pre-trained models and make them production-ready without expensive hardware.

GitHub - jina-ai/finetuner: Task-oriented finetuning for better embeddings on neural search

:dart: Task-oriented finetuning for better embeddings on neural search - GitHub - jina-ai/finetuner: Task-oriented finetuning for better embeddings on neural search

GitHubjina-ai

This release covers Finetuner version 0.7.2, including dependencies finetuner-api 0.5.2 and finetuner-core 0.12.7.

This release contains 2 new features, 4 refactorings, 1 bug fix, and 3 documentation improvements.

🆕 Features

Support learning rate scheduler (#679)

This PR add supports for the learning rate scheduler. The scheduler is used to adjust the learning rate during training. We support 6 learning rate schedulers: linear, cosine, cosine_with_restarts, polynomial, constant and constant_with_warmup.

When a scheduler is configured, the learning rate is by default adjusted after each batch. Alternatively, you can set scheduler_optons = {'scheduler_step': 'epoch'} to adjust the learning rate after each epoch instead.

You can use them by specifying their name in the scheduler attribute of the fit function.

run = finetuner.fit(
    ...,
    scheduler='linear',
    scheduler_options={'scheduler_step': 'batch'},
    ...
)

Support steps_per_interval in EvaluationCallback

When working with large datasets, you may want to perform evaluations multiple times during each epoch. This parameter allows to specify a number of batches after which an evaluation should be performed. If set to None, an evaluation is performed only at the end of each epoch.

run = finetuner.fit(
    ...,
    callbacks=[
        EvaluationCallback(
            query_data=...,
            index_data=...,
            steps_per_interval=3, # evaluate every 3 batches.
        ),
    ],
    ...
)

⚙ Refactoring

scheduler_step becomes part of scheduler_options (#679)

We removed the scheduler_step argument from the fit function, now it is part of the scheduler_options.

run = finetuner.fit(
    ...,
    scheduler='linear',
-   scheduler_step='batch',
+   scheduler_options={'scheduler_step': 'batch'},
    ...
)

Change the default epochs and batch_size in cloud.jina.ai

For Web UI users, we have reduced the default epochs from 10 to 5, and reduced the default batch_size from 128 to 64 to avoid out-of-memory errors from 3D-mesh fine-tuning.

Improve the user journey in cloud.jina.ai

Add more textual guidance on creating Finetuner runs in the Jina AI Cloud UI.

Remove duplicate query-document pairs in unlabeled CSVs (#678)

Finetuner now groups query-document pairs by their queries, thereby eliminating duplicate queries, when parsing CSV files. This leads to more effective fine-tuning.

🐞 Bug Fixes

Remove invalid argument from GeM pooler.

This PR removes the output_dim argument from the GeM pooler's forward function. You can use GeM pooler together with ArcFaceLoss to deliever better visual embedding quality.

run = finetuner.fit(
    ...,
    model_options = {
        ...
-       'output_dim': 512,
+       'pooler': 'GeM',
+       'pooler_options': {'p': 2.4, 'eps': 1e-5}
    }
)

📗 Documentation Improvements

Add a documentation section for GeM pooling (#684)

We have added a new section to our documentation which explains the pooling options in more detail.

Add a documentation page and notebook for ArcFaceLoss (#680)

We have added a new page to our documentation which demonstrates ArcFaceLoss on the Stanford Cars dataset.

Add a documentation section on creating query-document pairs as data (#678)

We have added a new section to our documentation explaining how to create training data made of query-document pairs instead of explicitly annotated and labeled data.

🤟 Contributors

We would like to thank all contributors to this release:

• Wang Bo (@bwanglzu)
• Louis Milliken (@LMMilliken)
• Michael Günther (@guenthermi)
• CatStark (@CatStark)
• George Mastrapas (@gmastrapas)
• Martin Matousek (@matousek-martin)
• Scott Martens (@scott-martens)