NLP inside your database - Query OpenAI’s GPT-3 from directly inside your database

@new_user_18120229380de99ff Thank you

@new_user_18120229380de99ff Thank you!

@new_user_18120229380de99ff thanks!

@new_user_18120229380de99ff Thanks for the support.

@chandre31 Great. Thanks for the clarification 👌🏿

@ed_forson Thank you

@ed_forson absolutely! please join our community on slack, Patricio Cerda from our team, has implemented the ability to finetune based on your data.

@ed_forson Yes, MindsDB allows users to train models against existing data in their database. The training process involves feeding the data into the MindsDB engine, which uses machine learning algorithms to learn the relationships between the input features and the target variables. During the training process, users have the option to modify various parameters, such as the learning rate, the number of trees, the maximum depth of the trees, and more, to optimize the performance of the model. In addition, MindsDB provides a user-friendly interface that makes it easy for users to train, test, and deploy their models. The interface allows users to visualize their data, set their target variables, and perform feature engineering tasks like normalizing and transforming their data. Once the model is trained, users can evaluate its performance using metrics like accuracy, precision, recall, and F1 score, and make further adjustments as needed. Overall, MindsDB is designed to be a flexible and user-friendly tool for building and deploying AI models and allows users to train models against their own data and adjust various parameters to optimize performance.

@usamaejaz thank you!! join our community on slack, we would be looking forward to hear about your discoveries.

@usamaejaz Thank you!

@usamaejaz Thanks!

@usamaejaz Thanks a lot. MindsDB community will always be there to help you in your ML journey.

@david_cardozo1 We are constantly integrating and receiving contributions for new integrations, if you are a developer I encourage you to create an integration (or many!) https://docs.mindsdb.com/contrib...

@david_cardozo1 Hey David! Thanks for your kind words. Here's a sheet that compares the default AutoML engine with other alternatives via the OpenML AutoML benchmarking suite: https://docs.google.com/spreadsh.... Hope it helps establish a rough idea on its performance! It's pretty old but we have a release procedure that helps us keep the accuracy, at the very least, so it should set a lower bound on what you should see for tabular regression and classification tasks.

@david_cardozo1 thanks

@david_cardozo1 THANK YOU! Firestore would be a great addition, lets chat about building one.

@david_cardozo1 MindsDB is constantly working on new integrations. You can easily open a Feature Request asking for the same and can even try writing the integration yourself if you want to. MindsDB does provide performance benchmarks for their battery-included models, which are available on their website. These benchmarks show the accuracy and speed of their models on a variety of datasets and tasks. Additionally, MindsDB's AutoML capabilities allow users to train custom models that can outperform the battery-included models for specific use cases. Overall, it's important to keep in mind that the performance of MindsDB's models will depend on many factors, including the size and complexity of the data, the type of task, and the computational resources available.

@ricardo_r_f_ @chandre31 especially this community is awesome!

@ricardo_r_f_ cool!

@ricardo_r_f_ nice!

@ricardo_r_f_ Thanks for sharing this.

@tomhuds We're close to shipping support for fine-tuned OpenAI completion models! After that, we'll look into doing the same for HuggingFace. As an aside, our MLOps features (particularly, model versioning and projects) enable you to effectively store, navigate and use all previous training runs for ML integrations that support the "retrain" and "adjust" commands with locally-stored models, which is quite convenient in use cases with long model lifecycles.

@tomhuds great question!

@tomhuds Stay tuned for further updates. We have much more exciting NLP stuff coming up pretty soon.

@martyna_slaw Thank you!

@martyna_slaw thanks

@martyna_slaw Thank you for all the love and support.

@david_fortunato1 wow thanks !

@david_fortunato1 Cool, thank you!

@david_fortunato1 Keep loving and supporting us.

@henryrearden I'd love to learn a bit more about your use case you can reach out to me on ac (at) mindsdb.com

@adam_carrigan @henryrearden looking forward to learn more about your use case, lets chat on our slack chat

@henryrearden Thanks for the appreciation. Feel free to join the community slack here- https://mindsdbcommunity.slack.com/ to discuss further.

@saasdesigner Thanks. We would love to hear your feedback about our UX.

@saasdesigner thanks

@saasdesigner Feel free to give it a try.