Actions364
- Continuous Activity Actions
- Dataset Actions
- Get Last Metric Values
- Get Metadata
- Get Schema
- Get Single Metric History
- List Datasets
- List Partitions
- Compute Metrics
- Create Dataset
- Create Managed Dataset
- Delete Data
- Delete Dataset
- Execute Tables Import
- Get Column Lineage
- Get Data
- Get Data - Alternative Version
- Get Dataset Settings
- Get Full Info
- List Tables
- List Tables Schemas
- Prepare Tables Import
- Run Checks
- Set Metadata
- Set Schema
- Synchronize Hive Metastore
- Update Dataset Settings
- Update From Hive Metastore
- API Service Actions
- Bundles Automation-Side Actions
- Bundles Design-Side Actions
- Connection Actions
- Dashboard Actions
- Data Collection Actions
- Data Quality Actions
- Compute Rules on Specific Partition
- Create Data Quality Rules Configuration
- Delete Rule
- Get Data Quality Project Current Status
- Get Data Quality Project Timeline
- Get Data Quality Rules Configuration
- Get Dataset Current Status
- Get Dataset Current Status per Partition
- Get Last Outcome on Specific Partition
- Get Last Rule Results
- Get Rule History
- Update Rule Configuration
- DSS Administration Actions
- Job Actions
- Library Actions
- Dataset Statistic Actions
- Discussion Actions
- Flow Documentation Actions
- Insight Actions
- Internal Metric Actions
- LLM Mesh Actions
- Machine Learning - Lab Actions
- Delete Visual Analysis
- Deploy Trained Model to Flow
- Download Model Documentation of Trained Model
- Generate Model Documentation From Custom Template
- Start Training ML Task
- Update User Metadata for Trained Model
- Update Visual Analysis
- Adjust Forecasting Parameters and Algorithm
- Compute Partial Dependencies of Trained Model
- Compute Subpopulation Analysis of Trained Model
- Create ML Task
- Create Visual Analysis
- Create Visual Analysis and ML Task
- Generate Model Documentation From Default Template
- Generate Model Documentation From File Template
- Get ML Task Settings
- Get ML Task Status
- Get Model Snippet
- Get Partial Dependencies of Trained Model
- Get Scoring Jar of Trained Model
- Get Scoring PMML of Trained Model
- Get Subpopulation Analysis of Trained Model
- Get Trained Model Details
- Get Visual Analysis
- List ML Tasks of Project
- List ML Tasks of Visual Analyses
- List Visual Analyses
- Reguess ML Task
- Machine Learning - Saved Model Actions
- Compute Partial Dependencies of Version
- Get Version Scoring PMML
- Get Version Snippet
- Import MLflow Version From File or Path
- List Saved Models
- List Versions
- Set Version Active
- Compute Subpopulation Analysis of Version
- Create Saved Model
- Delete Version
- Download Model Documentation of Version
- Evaluate MLflow Model Version
- Generate Model Documentation From Custom Template
- Generate Model Documentation From Default Template
- Generate Model Documentation From File Template
- Get MLflow Model Version Metadata
- Get Partial Dependencies of Version
- Get Saved Model
- Get Subpopulation Analysis of Version
- Get Version Details
- Get Version Scoring Jar
- Set Version User Meta
- Update Saved Model
- Long Task Actions
- Machine Learning - Experiment Tracking Actions
- Macro Actions
- Plugin Actions
- Download Plugin
- Fetch From Git Remote
- Get File Detail From Plugin
- Get Git Remote Info
- Get Plugin Settings
- Install Plugin From Git
- Install Plugin From Store
- List Files in Plugin
- List Git Branches
- List Plugin Usages
- Move File or Folder in Plugin
- Add Folder to Plugin
- Create Development Plugin
- Create Plugin Code Env
- Delete File From Plugin
- Delete Git Remote Info
- Delete Plugin
- Download File From Plugin
- Move Plugin to Dev Environment
- Pull From Git Remote
- Push to Git Remote
- Rename File or Folder in Plugin
- Reset to Local Head State
- Reset to Remote Head State
- Set Git Remote Info
- Set Plugin Settings
- Update Plugin Code Env
- Update Plugin From Git
- Update Plugin From Store
- Update Plugin From Zip Archive
- Upload File to Plugin
- Upload Plugin
- Project Deployer Actions
- Get Deployment Settings
- Get Deployment Status
- Create Deployment
- Create Infra
- Create Project
- Delete Bundle
- Delete Deployment
- Delete Infra
- Delete Project
- Get Deployment
- Get Deployment Governance Status
- Get Infra
- Get Infra Settings
- Get Project
- Get Project Settings
- Save Deployment Settings
- Save Infra Settings
- Save Project Settings
- Update Deployment
- Upload Bundle
- SQL Query Actions
- Wiki Actions
- Managed Folder Actions
- Meaning Actions
- Model Comparison Actions
- Notebook Actions
- Project Actions
- Project Folder Actions
- Recipe Actions
- Scenario Actions
- Security Actions
- Streaming Endpoint Actions
- Webapp Actions
- Workspace Actions
Overview
This node integrates with the Dataiku DSS API, enabling users to perform a wide range of operations on Dataiku DSS resources. Specifically for the LLM Mesh resource and the Perform Completions on a LLM operation, it allows performing text completions using large language models (LLMs) available within a Dataiku project.
Typical use cases include:
- Generating natural language completions or responses based on prompts.
- Automating content creation or augmentation workflows.
- Integrating advanced AI-driven text generation into data pipelines or applications managed in Dataiku.
For example, you might use this node to send a prompt to an LLM hosted in your Dataiku project and receive generated text completions that can be further processed or stored.
Properties
| Name | Meaning |
|---|---|
| Project Key | The key identifier of the Dataiku project where the LLM is hosted. |
| Request Body | JSON object containing the request payload for the completion operation. This typically includes the prompt and any parameters required by the LLM completion API. |
Note: The "Request Body" property expects a JSON structure defining the input for the LLM completion request, such as prompt text, model parameters, etc.
Output
The node outputs the response from the Dataiku DSS API call:
- The main output is a JSON object representing the completion results returned by the LLM service.
- If the response is a string, the node attempts to parse it as JSON; otherwise, it returns the raw text.
- For binary data (not typical for LLM completions), the node would prepare binary output, but this operation primarily deals with JSON responses.
The output JSON will contain the completion text and possibly additional metadata depending on the LLM's API response format.
Dependencies
- Requires valid Dataiku DSS API credentials, including:
- The URL of the Dataiku DSS server.
- A user API key for authentication.
- The node uses HTTP requests to communicate with the Dataiku DSS REST API.
- No other external dependencies are required beyond the configured Dataiku DSS API credentials.
Troubleshooting
- Missing Credentials Error: If the node throws an error about missing credentials, ensure that the Dataiku DSS API credentials are properly configured in n8n.
- Project Key Required: Many operations require the "Project Key" property. Make sure it is provided and correct.
- Invalid or Missing Parameters: The node validates required parameters per operation. Errors indicating missing parameters mean you need to supply those inputs.
- API Request Failures: Network issues, incorrect URLs, or invalid API keys can cause request failures. Check connectivity and credential validity.
- Response Parsing Errors: If the API returns non-JSON or unexpected responses, parsing may fail. Review the API response and adjust the request accordingly.
Links and References
This summary focuses on the LLM Mesh resource and the Perform Completions on a LLM operation, describing how the node constructs the API request, handles authentication, and processes the response for LLM completions within a Dataiku project.