- 2499 Views
- 1 replies
- 0 kudos
When would you use the Feature Store?
For example would you use a feature store on your raw data or what's is the granularity of the features in the store?
- 2499 Views
- 1 replies
- 0 kudos
- 0 kudos
I'll try to answer the broad question first, followed by the specific ones.When would you use the Feature Store?A Feature Store is primarily used to solve 2 challenges.(1) Discoverability and governance of featuresChallenge: In a large team or organi...
- 0 kudos
- 1394 Views
- 1 replies
- 0 kudos
- 1394 Views
- 1 replies
- 0 kudos
- 0 kudos
Yes.Please see Blog1: https://databricks.com/blog/2020/06/03/customer-lifetime-value-part-1-estimating-customer-lifetimes.htmlNotebook1:https://databricks.com/notebooks/CLV_Part_1_Customer_Lifetimes.htmlBlog2: https://databricks.com/blog/2020/06/17/c...
- 0 kudos
- 6294 Views
- 2 replies
- 0 kudos
Resolved! Can we delte Mlflow experiment
I am using ML flow and my need of the hour is to delete an experiment and want to create another experiment with same run.client = MlflowClient(tracking_uri=server) client.delete_experiment(1)This deletes the experiment, but when I run a new experim...
- 6294 Views
- 2 replies
- 0 kudos
- 0 kudos
SQL Database:This is more tricky, as there are dependencies that need to be deleted. I am using MySQL, and these commands work for me:USE mlflow_db; # the name of your database DELETE FROM experiment_tags WHERE experiment_id=ANY( SELECT experime...
- 0 kudos
- 5372 Views
- 1 replies
- 0 kudos
What's the best way to implement long term data versioning?
I'm a data scientist creating versioned ML models. For compliance reasons, I need to be able to replicate the training data for each model version. I've seen that you can version datasets by using delta, but the default retention period is around 30 ...
- 5372 Views
- 1 replies
- 0 kudos
- 0 kudos
Delta, as you mentioned has a feature to do time travel and by default, delta tables retain the commit history for 30 days. Operations on history of the table are parallel but will become more expensive as the log size increasesNow, in this case - s...
- 0 kudos
- 1804 Views
- 1 replies
- 0 kudos
- 1804 Views
- 1 replies
- 0 kudos
- 0 kudos
Yes.Please refer to our docshttps://docs.databricks.com/applications/machine-learning/manage-model-lifecycle/multiple-workspaces.html
- 0 kudos
- 2444 Views
- 1 replies
- 0 kudos
- 2444 Views
- 1 replies
- 0 kudos
- 0 kudos
Yes!You will have to pip install mlflowin your environment as a first step. For more details, see: https://docs.databricks.com/applications/mlflow/access-hosted-tracking-server.html
- 0 kudos
- 2464 Views
- 1 replies
- 0 kudos
Resolved! How is Databricks AutoML different than other AutoML products out there?
How does it provide a glass box view?
- 2464 Views
- 1 replies
- 0 kudos
- 0 kudos
Depending on which solution you use, GlassBox means that any interactive work you do via point & click, we automatically generate the code behind the scene and generate notebooks used for each experiment that was ran under the hood, in addition for a...
- 0 kudos
- 2423 Views
- 1 replies
- 0 kudos
What are the differences between Open Source and Hosted MLFlow?
We have been using open source MLflow, how will it benefit us to move to Databricks mlflow?
- 2423 Views
- 1 replies
- 0 kudos
- 0 kudos
Please see https://databricks.com/product/managed-mlflow
- 0 kudos
- 833 Views
- 0 replies
- 0 kudos
- 833 Views
- 0 replies
- 0 kudos
- 2292 Views
- 1 replies
- 1 kudos
What algorithms does Databricks AutoML use?
AutoML presumably tries a few different algorithms while hyperparameter searching. What model types are considered?
- 2292 Views
- 1 replies
- 1 kudos
- 1 kudos
At the moment, it's really just xgboost, and sklearn implemenations like random forests, logistic regression, and linear regression as applicable. More possibilities are coming.
- 1 kudos
- 2080 Views
- 1 replies
- 0 kudos
How can I use Non- Spark related libraries like spacy with Databricks and Spark
I have an NLP application that I build on my local machine using spacy and pandas, but now I would like to scale my application to a large production dataset and utilize the benefits of sparks distributed compute. How do I import and utilize a librar...
- 2080 Views
- 1 replies
- 0 kudos
- 0 kudos
It depends on what you mean, but if you're just trying to (say) tokenize and process data with spacy in parallel, then that's trivial. Write a 'pandas UDF' function that expresses how you want to transform data using spacy, in terms of a pandas DataF...
- 0 kudos
- 3387 Views
- 1 replies
- 0 kudos
Resolved! Is there documentation about the feature store API and how it's architected under the hood?
- 3387 Views
- 1 replies
- 0 kudos
- 0 kudos
I don't think we have a lot of internal docs, just high-level explanations like https://databricks.com/blog/2021/05/27/databricks-announces-the-first-feature-store-integrated-with-delta-lake-and-mlflow.htmlHowever I don't think there's much to it. Th...
- 0 kudos
- 2647 Views
- 1 replies
- 0 kudos
- 2647 Views
- 1 replies
- 0 kudos
- 0 kudos
The feature store has both online / offline components. The offline feature store is used for feature discovery, model training, and batch inference and is backed by Delta tables. You could read/write to offline store from Databricks clusters that...
- 0 kudos
- 3458 Views
- 1 replies
- 1 kudos
What are best NLP libraries to use with Spark
Best NLP APIs to use with Spark which gives better performance
- 3458 Views
- 1 replies
- 1 kudos
- 1 kudos
By far the most popular and comprehensive library, to my knowledge, for Spark-native distributed NLP, is spark-nlp from John Snow Labs. https://nlp.johnsnowlabs.com/ It is open source (but with commercial support options) and has a whole lot of funct...
- 1 kudos
- 4542 Views
- 1 replies
- 0 kudos
- 4542 Views
- 1 replies
- 0 kudos
- 0 kudos
These terms are borrowed from scikit-learn, and the idea is the same. A transformer is just a component of a pipeline that transforms the data in some way. An estimator is also a transfomer, but one that additionally needs to be 'fit' on data before ...
- 0 kudos
