Product area

Experiments

MLflow Experiments: tracking runs, parameters, metrics, and artifacts.

Concepts

10 in this area
  1. 1MLflow tracking on DatabricksMLflow Tracking is the piece of MLflow that records what happened during a training run: which parameters you used, which metrics came out, and which files it produced (model weights, plots, a requirements.txt). Every run belongs to an experiment, a named…
  2. 2MLflow 3 for modelsMLflow 3 is the version of MLflow that the current Databricks machine learning and generative AI tooling is built on. Experiments and runs still mean what they always meant (see mlflow tracking), but the thing you actually care about has been promoted: a…
  3. 3MLflow Tracing for GenAI applicationsMLflow Tracing is the observability layer for generative AI code. One call into your application produces one trace: the record of everything that happened between the request arriving and the answer going out. A trace is a tree of spans, and a span is one…
  4. 4Evaluation datasets for generative AIAn evaluation dataset is the fixed set of examples an application is scored against. Each record has inputs, a dictionary holding whatever the application takes (a question, a conversation, some context), and optionally expectations, a dictionary holding what…
  5. 5Human feedback on generative AI outputHuman feedback in MLflow is stored as an assessment attached to a trace, or to a single span inside one. There are two kinds, and the distinction runs through everything else on this page:
  6. 6Prompt registryThe prompt registry stores a prompt template as an object in Unity Catalog instead of as a string in your source code. A prompt has a three-level name such as main.genai.supportsummary, immutable versions numbered automatically as you register new text…
  7. 7Production monitoring for GenAI appsProduction monitoring runs the scorers from agent evaluation against traffic instead of against a dataset. You take a scorer you already trust, register it against the MLflow experiment your app logs to, and start it with a sampling rate. From then on a…
  8. 8Training and tuning a classic modelTraining a classic model on Databricks is mostly ordinary Python. The machine learning runtime ships scikit-learn, XGBoost, PyTorch and TensorFlow already installed, and a notebook on a single node runs them the way a laptop would, with more memory.
  9. 9AutoMLAutoML takes a table, a target column and a problem type, then trains a lot of models and tells you which did best. It covers classification, regression and forecasting.
  10. 10AI Runtime and serverless GPU computeAI Runtime is serverless GPU compute for training and fine-tuning. You ask for a number of accelerators of a given type, Databricks provisions them, runs your Python on them, and releases them when the work finishes. There is no cluster to size, no driver to…

News

No news for this area.

Resources

AI RuntimeDocumentation · DatabricksFreeDistributed training in notebooksDocumentation · DatabricksFreeAI Runtime CLIDocumentation · DatabricksFreeInstall the AI Runtime CLIDocumentation · DatabricksFreeAI Runtime CLI command referenceDocumentation · DatabricksFreeWorkload YAML referenceDocumentation · DatabricksFreeFoundation Model Fine-tuning (EoL)Documentation · DatabricksFreeDatabricks-hosted foundation models available in Foundation Model APIsDocumentation · DatabricksFreeWhat is AutoML?Documentation · DatabricksFreeBuilding MLflow evaluation datasetsDocumentation · DatabricksFreeEvaluation dataset referenceDocumentation · DatabricksFreeSynthesize evaluation setsDocumentation · DatabricksFreeCreate and manage labeling sessionsDocumentation · DatabricksFreeMonitor GenAI apps in productionDocumentation · DatabricksFreeScorer lifecycle management API referenceDocumentation · DatabricksFreeTrace agents deployed on DatabricksDocumentation · DatabricksFreeHuman feedback in MLflowDocumentation · DatabricksFreeLabel during developmentDocumentation · DatabricksFreeCollect structured human feedback with review queuesDocumentation · DatabricksFreeCollect feedback and expectations by labeling existing tracesDocumentation · DatabricksFreeCollect user feedbackDocumentation · DatabricksFreeGet started with MLflow 3 for modelsDocumentation · DatabricksFreeTrack and compare models using MLflow Logged ModelsDocumentation · DatabricksFreeMigrate to MLflow 3 from Agent Evaluation: Quick referenceDocumentation · DatabricksFreeDatabricks Runtime 17.3 LTS for Machine LearningDocumentation · DatabricksFreePrompt RegistryDocumentation · DatabricksFreeMLflow Prompt Optimization (beta)Documentation · DatabricksFreeMLflow Tracing - GenAI observabilityDocumentation · DatabricksFreeAdd traces to applications: automatic and manual tracingDocumentation · DatabricksFreeAutomatic tracingDocumentation · DatabricksFreeFunction decoratorsDocumentation · DatabricksFreeSpan tracing with context managersDocumentation · DatabricksFreeSpan conceptsDocumentation · DatabricksFreeAccess trace dataDocumentation · DatabricksFreeStore OpenTelemetry traces in Unity CatalogDocumentation · DatabricksFreeQuery model APIs (model services)Documentation · DatabricksFreeHyperparameter tuningDocumentation · DatabricksFreeHyperparameter tuning with OptunaDocumentation · DatabricksFreeTrain AI and ML modelsDocumentation · DatabricksFreeCreate and edit promptsDocumentation · DatabricksFreeUse prompts in deployed applicationsDocumentation · DatabricksFreeEvaluate and compare prompt versionsDocumentation · DatabricksFreeMLflow 3.0: AI and MLOps on DatabricksVideo · Databricks · checked 10 Sept 2026FreeTrack model development using MLflowDocumentation · Databricks · checked 10 Sept 2026Free