Taalas API

openapi: 3.1.0
info:
  title: Taalas API
  description: Taalas API interface for running inference against the HC1 harcore model silicon.
  version: 1.0.0
paths:
  /health:
    get:
      tags:
      - Monitoring
      summary: Health
      description: "Endpoint to check the health status of the server.\n\nReturns:\n    dict: A dictionary\
        \ containing the following keys:\n        - status (str): The health status of the server, always\
        \ \"healthy\".\n        - queue_size (int): The number of items in the queue managed by `queue_manager`.\n\
        \        - current_adapter (str): The name of the currently loaded LoRA model, or \"none\" if\
        \ no model is loaded."
      operationId: health_health_get
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HealthResponse'
  /models:
    get:
      tags:
      - Models
      summary: Models
      description: Endpoint to get the model info of server.
      operationId: models_models_get
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema: {}
  /generate:
    post:
      tags:
      - Inference
      summary: Generate
      description: Generate text based on the input prompt.
      operationId: generate_generate_post
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/GenerateRequest'
        required: true
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema: {}
        '422':
          description: Validation Error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HTTPValidationError'
components:
  schemas:
    GenerateRequest:
      properties:
        prompt:
          items:
            type: object
          type: array
          title: Prompt
          description: The input prompt for the language model to generate text from.
        min_output_tokens:
          type: integer
          minimum: 0.0
          title: Min Output Tokens
          description: Minimum number of tokens to output.
          default: 0
        max_output_tokens:
          type: integer
          minimum: 0.0
          title: Max Output Tokens
          description: Maximum number of tokens to output.
          default: 0
        topk:
          type: integer
          maximum: 8.0
          minimum: 1.0
          title: Topk
          description: Choose top k tokens.
          default: 1
        model:
          type: string
          title: Model
          description: The name of the model or adapter to use for generation.
          default: llama3.1-8B
        stream:
          anyOf:
          - type: boolean
          - type: 'null'
          title: Stream
          description: Whether to stream the response back incrementally.
          default: false
        tools:
          anyOf:
          - items:
              type: object
            type: array
          - type: 'null'
          title: Tools
          description: List of tools given to model.
        tool_choice:
          anyOf:
          - type: string
          - type: 'null'
          title: Tool Choice
          description: Controls which (if any) tool is called by the model.
        logprobs:
          type: boolean
          title: Logprobs
          description: Whether to return log probabilities of the output tokens or not.
          default: false
        top_logprobs:
          type: integer
          maximum: 20.0
          minimum: 0.0
          title: Top Logprobs
          description: Return output tokens with top probabilities.
          default: 0
        system_prompt_version:
          type: string
          title: System Prompt Version
          description: Which system prompt to use in the backend.
          default: offline
        tool_name:
          type: string
          title: Tool Name
          description: Which tool was selected by router LLM model.
          default: ''
        context_for_tool:
          type: string
          title: Context For Tool
          description: Context most relavant to selected tool.
          default: ''
        guided_json:
          anyOf:
          - type: object
          - type: 'null'
          title: Guided Json
          description: A JSON schema for guided generation.
        guided_regex:
          anyOf:
          - type: string
          - type: 'null'
          title: Guided Regex
          description: A regex pattern for guided generation.
        response_format:
          anyOf:
          - type: object
          - type: 'null'
          title: Response Format
          description: For JSON object output.
        metadata:
          anyOf:
          - type: object
          - type: 'null'
          title: Metadata
          description: Used to propagate number of prefill and decode tokens.
        echo:
          type: boolean
          title: Echo
          description: Echo back the prompt in addition to the completion.
          default: false
        stream_options:
          anyOf:
          - $ref: '#/components/schemas/StreamOptions'
          - type: 'null'
          description: 'If stream: true, this will cause the server to send back usage stats as a final
            chunk.'
        stop:
          anyOf:
          - items:
              type: string
            type: array
          - type: string
          - type: 'null'
          title: Stop
          description: Up to 4 sequences where the API will stop generating further tokens. The returned
            text will not contain the stop sequence.
        temperature:
          anyOf:
          - type: number
            maximum: 2.0
            minimum: 0.0
          - type: 'null'
          title: Temperature
          description: What sampling temperature to use, between 0 and 2.
          default: 1
        ignore_eos:
          type: boolean
          title: Ignore Eos
          description: Ignore end of sequence (stop_id) and continue generation until output limit is
            hit.
          default: false
      additionalProperties: true
      type: object
      required:
      - prompt
      title: GenerateRequest
    HTTPValidationError:
      properties:
        detail:
          items:
            $ref: '#/components/schemas/ValidationError'
          type: array
          title: Detail
      type: object
      title: HTTPValidationError
    HealthResponse:
      properties:
        status:
          type: string
          title: Status
          description: Health status of the server.
        queue_size:
          type: integer
          title: Queue Size
          description: Number of items in the processing queue.
        current_adapter:
          type: string
          title: Current Adapter
          description: Name of the currently loaded LoRA adapter.
      type: object
      required:
      - status
      - queue_size
      - current_adapter
      title: HealthResponse
    StreamOptions:
      properties:
        include_usage:
          anyOf:
          - type: boolean
          - type: 'null'
          title: Include Usage
          default: false
        include_obfuscation:
          anyOf:
          - type: boolean
          - type: 'null'
          title: Include Obfuscation
          default: false
      type: object
      title: StreamOptions
    ValidationError:
      properties:
        loc:
          items:
            anyOf:
            - type: string
            - type: integer
          type: array
          title: Location
        msg:
          type: string
          title: Message
        type:
          type: string
          title: Error Type
      type: object
      required:
      - loc
      - msg
      - type
      title: ValidationError