> ## Documentation Index > Fetch the complete documentation index at: https://docs.mathfi.ai/llms.txt > Use this file to discover all available pages before exploring further. # Get training job detail > Get training job detail by key ## OpenAPI ````yaml api-reference/openapi.yaml GET /api/v1/training/{trainingJobKey} openapi: 3.0.1 info: title: MathFi.ai API description: > The MathFi.ai REST API runs the two products end to end. **Feature Refinery** takes your labelled data and works out which columns actually carry the signal. It returns a reduced train and test pair, a per-column keep/remove decision, and the performance target the refined data reached. **Model Crucible** takes a dataset, trains every algorithm family against it at once, ranks the results on held-out data by fewest wrong decisions, and keeps the best three as versions of one model. You pick which one predictions run against. Everything is asynchronous: you create a thing, upload to a signed URL, start it, then poll until the state is terminal. Nothing streams and nothing blocks. contact: name: MathFi.ai url: https://mathfi.ai email: support@mathfi.ai license: name: MathFi.ai url: https://mathfi.ai version: 1.0.0 servers: - url: https://{tenant}-api.mathfi.ai description: >- Your tenant's API. Each customer has their own, so the host varies. Replace {tenant} with the name issued when your tenant was created. variables: tenant: default: your-tenant description: The tenant name issued to you. security: - BearerAuth: [] tags: - name: Authentication description: Exchange credentials for a bearer token - name: Feature Refinery description: Reduce a dataset to the columns that earn their place - name: Datasets description: >- Prepare labelled data for training, from uploads or from a finished refinement - name: Training description: Run the Crucible against a dataset and choose the champion model - name: Models description: Trained models and their versions - name: Predictions description: Score unlabelled data against a champion model paths: /api/v1/training/{trainingJobKey}: get: tags: - Training summary: Get training job detail description: Get training job detail by key operationId: getTrainingDetail parameters: - name: trainingJobKey in: path required: true schema: type: string format: uuid responses: '200': description: Training job detail successful response content: application/json: schema: $ref: '#/components/schemas/TrainingJobOutput' examples: Successful training job detail response: summary: Successful training job detail response value: trainingJobKey: 3a1a9a83-d549-4f25-b7ed-8174e0c955de datasetKey: 2b0a9a83-d549-4f25-b7ed-8174e0c955cd status: PENDING scalingFactor: 19 targetPerformance: 0.85 '400': description: Invalid training detail request '500': description: Internal server error components: schemas: TrainingJobOutput: required: - datasetKey - status - targetPerformance - trainingJobKey type: object properties: trainingJobKey: type: string format: uuid datasetKey: type: string format: uuid status: $ref: '#/components/schemas/TrainingJobStatus' createdOn: description: >- When the run was created, as an ISO-8601 UTC instant. The listing carries this too; it is here so a client that opened a run directly does not have to go back to the list for it. type: string nullable: true scalingFactor: type: integer format: int32 nullable: true targetPerformance: type: number format: double achievedPerformance: type: number format: double nullable: true trainingPerformance: type: number format: double testPerformance: type: number format: double modelKey: type: string format: uuid nullable: true winnerAlgorithm: type: string nullable: true description: Alias of the winning algorithm. Present only on COMPLETED jobs. algorithmPerformances: type: array description: >- Per-algorithm latest-in-time performance at job termination. Present on all terminal statuses. items: $ref: '#/components/schemas/TrainingJobProgressItem' performance: $ref: '#/components/schemas/TrainingPerformance' TrainingJobStatus: type: string description: > Status of training: * `PENDING` - The training job has been created and is pending execution * `RUNNING` - The training job is currently executing and progressing * `COMPLETED` - The training job has completed execution successfully, reaching the target performance. A model has been generated. * `TIMED_OUT` - The training job could not reach the target performance in the configured * `CANCELLED` - The training job has been cancelled by the user * `NOT_COMPLETED` - The training job has stalled without making progress in a specified timeframe * `FAILED` - The training job has failed due to an error enum: - PENDING - RUNNING - COMPLETED - TIMED_OUT - CANCELLED - FAILED - NOT_COMPLETED TrainingJobProgressItem: type: object properties: trainingJobKey: type: string format: uuid attemptNumber: description: >- Which of the job's attempts this is, counting from one. Present where a job runs many attempts rather than one job per algorithm — the attempts have no key of their own. type: integer nullable: true algorithm: type: string status: $ref: '#/components/schemas/TrainingJobStatus' latestPerformance: type: number format: double recentPerformances: type: array items: type: number format: double TrainingPerformance: description: >- What a finished run produced: the models it offers, which of them predictions run against, and where they were published. Null on a run of the older pipeline, which offers one model and no choice. type: object properties: candidates: description: The models this run offers, best first. Position is the ranking. type: array items: $ref: '#/components/schemas/ModelCandidate' champion: $ref: '#/components/schemas/ChampionSelectionOutput' achievedPerformance: description: >- What this run achieved: the accuracy of the model predictions currently run against. It follows the champion, so re-pointing the choice changes it. Absent while nothing has been offered, and on a champion whose accuracy the jars could not define. type: number format: double nullable: true modelKey: type: string format: uuid nullable: true ModelCandidate: description: One model a finished run offers, and how it scored on the test data. type: object properties: attemptNumber: type: integer algorithm: type: string version: type: string falseDecisions: description: >- How many decisions it got wrong. Fewer is better; this is what the ranking is on. type: integer accuracy: type: number format: double nullable: true f1Score: type: number format: double nullable: true precision: type: number format: double nullable: true recall: type: number format: double nullable: true ChampionSelectionOutput: description: Which candidate predictions run against, and how that came to be. type: object properties: attemptNumber: type: integer version: type: string selectedBy: type: string enum: - AUTOMATIC - USER observations: description: >- Why this candidate — the run's own reason, or the user's own words over it. type: string nullable: true securitySchemes: BearerAuth: type: http scheme: bearer bearerFormat: JWT ````