Topics¶
Topics connect tutorials, procedures, contracts, and design explanations that concern the same subsystem.
Vocabulary¶
alchemy: image interrogation and post-processing requests.api-models: typed request and response objects for Horde APIs.architecture: package boundaries, abstraction layers, and extension points.backends: normalization between Horde jobs and inference backends.chaining: dependency graphs that compose worker operations.clients: synchronous, asynchronous, simple, manual, and session clients.configuration: deployment and runtime configuration contracts.contributing: repository conventions and maintenance workflows.dispatch: selecting and preparing work for a backend.errors: failure types, cleanup behavior, and troubleshooting.generation: generation parameters, requests, progress, and results.images: image generation parameters, files, and utilities.model-reference: integration with the Horde model reference.ratings: ratings API clients and models.requests: request submission, polling, responses, and cleanup.safety: content-safety and localization utilities.text: text generation requests and parameters.utilities: logging, localization, image, and support helpers.workers: worker jobs, generations, state, and lifecycle.
By topic¶
alchemy¶
- How-to: Run an alchemy operation
- How-to: Add an alchemy form (planned)
- Reference: Generation parameters and features
api-models¶
- Explanation: Typed request and response design
- How-to: Add an alchemy form (planned)
- How-to: Update endpoint models (planned)
- Reference: Request and response contracts
- Reference: AI Horde endpoint map
architecture¶
- Explanation: Client layers and ownership
- Explanation: Typed request and response design
- Explanation: Generation parameters and templates
- Explanation: Worker jobs, generations, and state
- Explanation: Package architecture
- Explanation: Backend normalization
- Explanation: Operation chaining (planned)
- Explanation: Ratings integration (planned)
- Explanation: Model-reference integration design (planned)
- How-to: Implement a backend value mapper
- Reference: Package and code map
- Reference: SDK glossary
- Reference: Contributor conventions
backends¶
- Explanation: Worker jobs, generations, and state
- Explanation: Backend normalization
- Explanation: Dispatch normalization
- How-to: Implement a backend value mapper
- How-to: Drive the worker generation state machine
- How-to: Implement a dispatch adapter (planned)
- Reference: Worker generation transitions
- Reference: Backend parsing contracts (planned)
- Reference: Dispatch interfaces (planned)
- Tutorial: Build your first worker integration (planned)
chaining¶
- Explanation: Operation chaining (planned)
- How-to: Build a chained worker flow (planned)
- Reference: Chaining interfaces (planned)
clients¶
- Explanation: Client layers and ownership
- How-to: Generate images
- How-to: Generate text
- How-to: Use the asynchronous client
- How-to: Estimate request cost
- How-to: Run an alchemy operation
- How-to: Clean up manual requests
- How-to: Use the ratings API (planned)
- Reference: Client behavior and errors
- Reference: Troubleshooting
- Reference: Ratings API reference (planned)
- Tutorial: Create and verify your first image
configuration¶
- Reference: Deployment configuration (planned)
contributing¶
- Explanation: Package architecture
- How-to: Update endpoint models (planned)
- Reference: Package and code map
- Reference: Contributor conventions
dispatch¶
- Explanation: Backend normalization
- Explanation: Dispatch normalization
- How-to: Drive the worker generation state machine
- How-to: Add an alchemy form (planned)
- How-to: Implement a dispatch adapter (planned)
- Reference: Dispatch interfaces (planned)
- Tutorial: Build your first worker integration (planned)
errors¶
- How-to: Clean up manual requests
- Reference: Client behavior and errors
- Reference: Request and response contracts
- Reference: Troubleshooting
- Reference: Logging reference (planned)
generation¶
- Explanation: Generation parameters and templates
- Explanation: Worker jobs, generations, and state
- Explanation: Backend normalization
- Explanation: Model-reference integration design (planned)
- How-to: Generate images
- How-to: Generate text
- How-to: Use the asynchronous client
- How-to: Estimate request cost
- How-to: Implement a backend value mapper
- How-to: Drive the worker generation state machine
- Reference: Generation parameters and features
- Reference: Worker generation transitions
- Reference: SDK glossary
- Reference: Troubleshooting
- Reference: Backend parsing contracts (planned)
- Reference: Model-reference integration (planned)
- Tutorial: Create and verify your first image
images¶
- Explanation: Generation parameters and templates
- How-to: Generate images
- How-to: Use the asynchronous client
- How-to: Run an alchemy operation
- Reference: Generation parameters and features
- Reference: Image utilities (planned)
- Tutorial: Create and verify your first image
model-reference¶
- Explanation: Model-reference integration design (planned)
- Reference: Model-reference integration (planned)
ratings¶
- Explanation: Ratings integration (planned)
- How-to: Use the ratings API (planned)
- Reference: Ratings API reference (planned)
requests¶
- Explanation: Client layers and ownership
- Explanation: Typed request and response design
- How-to: Generate images
- How-to: Generate text
- How-to: Use the asynchronous client
- How-to: Estimate request cost
- How-to: Run an alchemy operation
- How-to: Clean up manual requests
- How-to: Update endpoint models (planned)
- How-to: Use the ratings API (planned)
- Reference: Client behavior and errors
- Reference: Request and response contracts
- Reference: SDK glossary
- Reference: AI Horde endpoint map
- Reference: Ratings API reference (planned)
- Tutorial: Create and verify your first image
safety¶
- Reference: Safety and localization utilities (planned)
text¶
- Explanation: Generation parameters and templates
- How-to: Generate text
- Reference: Generation parameters and features
utilities¶
- Reference: Logging reference (planned)
- Reference: Safety and localization utilities (planned)
- Reference: Image utilities (planned)
workers¶
- Explanation: Worker jobs, generations, and state
- Explanation: Dispatch normalization
- Explanation: Operation chaining (planned)
- How-to: Drive the worker generation state machine
- How-to: Implement a dispatch adapter (planned)
- How-to: Build a chained worker flow (planned)
- Reference: Worker generation transitions
- Reference: SDK glossary
- Reference: Troubleshooting
- Reference: Dispatch interfaces (planned)
- Reference: Chaining interfaces (planned)
- Reference: Deployment configuration (planned)
- Tutorial: Build your first worker integration (planned)