Lamini memory rag API

openapi: 3.2.0
info:
  title: Lamini memory rag API
  description: LLM inference and tuning for the enterprise. Factual LLMs. Classifier and embeddings. Deployed anywhere in 10 min.
  version: 1.13.6
security:
- APIKeyAuth: []
tags:
- name: memory rag
paths:
  /alpha/memory-rag/add-index:
    post:
      tags:
      - memory rag
      summary: Append to an existing an Memory RAG index
      description: Append to an existing an Memory RAG index with the provided prompt.
      operationId: memory_rag_add_index_post
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/MemoryRagAddIndexPayload'
            example:
              prompt: Content to input into the index.
              job_id: '1'
        required: true
      responses:
        '200':
          description: Successful response
          content:
            application/json:
              schema: {}
              example:
                job_id: 1
                status: success
        '422':
          description: Validation Error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HTTPValidationError'
      security:
      - API Key Auth: []
      parameters:
      - name: Authorization
        in: header
        required: true
        schema:
          type: string
        description: API key for authentication
  /alpha/memory-rag/completions:
    post:
      tags:
      - memory rag
      summary: Generate Memory RAG completions for prompts
      description: Generate a Memory RAG inference response to a prompt, optionally with a specified JSON output structure, with a `finish_reason` field in the response indicating the end state for each inference request.
      operationId: memory_rag_completions_post
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/MemoryRagCompletionsPayload'
        required: true
      responses:
        '200':
          description: Successful response
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/CompletionsWithFinishReasonResponse'
              example:
                outputs:
                - text: This is one inference output.
                - text: This is another output.
                finish_reason: stop
        '422':
          description: Validation Error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HTTPValidationError'
      security:
      - API Key Auth: []
      parameters:
      - name: Authorization
        in: header
        required: true
        schema:
          type: string
        description: API key for authentication
  /alpha/memory-rag/status:
    get:
      tags:
      - memory rag
      summary: Get the status of a Memory RAG training job
      description: Get the status of the Memory RAG training job. Results in either 'created', 'running', 'failed', or 'completed'.
      operationId: memory_rag_job_status
      parameters:
      - required: true
        schema:
          type: string
          title: Job Id
        name: job_id
        in: path
      - name: Authorization
        in: header
        required: true
        schema:
          type: string
        description: API key for authentication
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema: {}
            example:
              outputs:
                status: completed
        '422':
          description: Validation Error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HTTPValidationError'
      security:
      - API Key Auth: []
  /alpha/memory-rag/suggestions:
    post:
      tags:
      - memory rag
      summary: Retrieve a sample from a Memory RAG Index
      description: Retrieve a sample of 5 examples from a Memory RAG Index for the provided job id.
      operationId: memory_rag_suggestions_post
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/MemoryRagSuggestionsPayload'
        required: true
      responses:
        '200':
          description: Successful response
          content:
            application/json:
              schema: {}
              example:
              - Question 1: Example Question 1
                Answer 1: Example Answer 1
              - Question 2: Example Question 2
                Answer 2: Example Answer 2
        '422':
          description: Validation Error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HTTPValidationError'
      security:
      - API Key Auth: []
      parameters:
      - name: Authorization
        in: header
        required: true
        schema:
          type: string
        description: API key for authentication
  /alpha/memory-rag/train:
    post:
      tags:
      - memory rag
      summary: Start a Memory RAG job using the provided files
      description: Start a Memory RAG job using the provided files, resulting in a Memory RAG Index available for Completion calls.
      operationId: memory_rag_train_post
      requestBody:
        content:
          application/json:
            schema: {}
            example:
              files: '{path/to/document}.pdf'
              model_name: meta-llama/Llama-3.1-8B-Instruct
        required: true
      responses:
        '200':
          description: Successful response
          content:
            application/json:
              schema: {}
              example:
                job_id: 1
                status: CREATED
        '422':
          description: Validation Error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HTTPValidationError'
      security:
      - API Key Auth: []
      parameters:
      - name: Authorization
        in: header
        required: true
        schema:
          type: string
        description: API key for authentication
components:
  schemas:
    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
    HTTPValidationError:
      properties:
        detail:
          items:
            $ref: '#/components/schemas/ValidationError'
          type: array
          title: Detail
      type: object
      title: HTTPValidationError
    FinishReason:
      type: string
      enum:
      - stop
      - length
      - timeout
      title: FinishReason
      description: 'Indicates the ending state of the inference request.


        `stop`: The model completed the response with an EOS or equivalent token. This means the model finished generating the response.


        `length`: The response reached the maximum number of tokens as set by max_new_tokens or other token limit.


        `timeout`: The time taken to generate the response exceeded the configured Lamini Platform inference timeout. See [Performance](https://docs.lamini.ai/inference/performance/) for how to use this information to debug problems and improve performance.'
    CompletionsWithFinishReasonResponse:
      properties:
        outputs:
          items:
            additionalProperties:
              type: string
            type: object
          type: array
          title: Outputs
          description: The inference outputs for the given prompt(s).
        finish_reason:
          items:
            $ref: '#/components/schemas/FinishReason'
          type: array
          title: Finish Reason
          description: The final state for each inference request.
      type: object
      required:
      - outputs
      - finish_reason
      title: CompletionsWithFinishReasonResponse
  securitySchemes:
    APIKeyAuth:
      type: apiKey
      in: header
      name: Authorization
      description: Enter your API token. Get your API key from http://localhost:5001/account
x-provenance:
  method: harvested
  publisher: Lamini
  source: https://docs.lamini.ai/openapi.json
  note: split by refine-openapis from the provider's own published document, verified by fetch (openapi/_original/*.fetch.json);
    stamped by stamp-openapi-provenance.py (roadmap#793)