Some products and features are in the process of being renamed. Generative playbook and flow features are also being migrated to a single consolidated console. See the details.
Stay organized with collections
Save and categorize content based on your preferences.
When your agent is trained,
Conversational Agents (Dialogflow CX) uses your training data to build machine learning models
specifically for your agent.
Each flow of your agent has its own model.
This training data primarily consists of intents, intent training phrases,
and entities referenced in a flow;
which are effectively used as machine learning data labels.
However, agent models are built using
parameter prompt responses, state handlers,
agent settings, and many other pieces of data associated with your agent.
Whenever you change your agent,
you should ensure that the agent is trained before attempting to use it.
Depending on your agent settings,
training may occur automatically or manually.
When training is initiated by the console, the
task indicator
provides status for the tasks.
Draft flow automatic training
ML settings
lets you select either a Standard NLU or Advanced NLU NLU type.
A setting to enable or disable automatic training is available.
If you have disabled automatic training,
you must execute manual training
after you have completed updates.
Draft flow manual training
To manually train a flow from the console,
click the Train NLU button in the
ML settings.
To manually train a flow with the API,
call the train method for the Flow type.
Select a protocol and version for the Flow reference:
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Hard to understand","hardToUnderstand","thumb-down"],["Incorrect information or sample code","incorrectInformationOrSampleCode","thumb-down"],["Missing the information/samples I need","missingTheInformationSamplesINeed","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2025-08-07 UTC."],[[["\u003cp\u003eDialogflow CX trains machine learning models for each agent flow using intents, training phrases, and entities, among other data.\u003c/p\u003e\n"],["\u003cp\u003eIt's crucial to train the agent after making any changes to ensure it functions properly, which can be configured to occur automatically or manually.\u003c/p\u003e\n"],["\u003cp\u003eManual training of a flow can be initiated through the "Train NLU" button in the console's ML settings or via the API's \u003ccode\u003etrain\u003c/code\u003e method for the \u003ccode\u003eFlow\u003c/code\u003e type.\u003c/p\u003e\n"],["\u003cp\u003eNew flow versions are automatically trained upon creation, either through the console's "Create" button in the Version panel or via the API's \u003ccode\u003ecreate\u003c/code\u003e method for the \u003ccode\u003eVersion\u003c/code\u003e type.\u003c/p\u003e\n"]]],[],null,["# Training\n\nWhen your agent is trained,\nConversational Agents (Dialogflow CX) uses your training data to build machine learning models\nspecifically for your agent.\nEach flow of your agent has its own model.\nThis training data primarily consists of intents, intent training phrases,\nand entities referenced in a flow;\nwhich are effectively used as machine learning data labels.\nHowever, agent models are built using\nparameter prompt responses, state handlers,\nagent settings, and many other pieces of data associated with your agent.\n\nWhenever you change your agent,\nyou should ensure that the agent is trained before attempting to use it.\nDepending on your agent settings,\ntraining may occur automatically or manually.\n\nWhen training is initiated by the console, the\n[task indicator](/dialogflow/cx/docs/concept/console#task)\nprovides status for the tasks.\n\nDraft flow automatic training\n-----------------------------\n\n[ML settings](/dialogflow/cx/docs/concept/agent-settings#settings-ml)\nlets you select either a **Standard NLU** or **Advanced NLU** NLU type.\nA setting to enable or disable automatic training is available.\n\nIf you have disabled automatic training,\nyou must execute [manual training](#draft-manual)\nafter you have completed updates.\n\nDraft flow manual training\n--------------------------\n\nTo manually train a flow from the console,\nclick the **Train NLU** button in the\n[ML settings](/dialogflow/cx/docs/concept/agent-settings#settings-ml).\n\nTo manually train a flow with the API,\ncall the `train` method for the `Flow` type.\n\n\nGo to the Flow API reference \n**Select a protocol and version for the Flow reference:**\n\nClose\n\n\u003cbr /\u003e\n\nFlow version automatic training\n-------------------------------\n\nWhenever a new\n[flow version](/dialogflow/cx/docs/concept/version)\nis created, the new flow version is automatically trained.\n\nTo create a new flow version from the console,\nclick the **Create** button on the\n[Version panel](/dialogflow/cx/docs/concept/version#manage-versions).\n\nTo create a new flow version with the API,\ncall the `create` method for the `Version` type.\n\n\nGo to the Version API reference \n**Select a protocol and version for the Version reference:**\n\nClose\n\n\u003cbr /\u003e"]]
