{ "$schema": "https://json-schema.org/draft/2020-12/schema", "$id": "https://docs.mistral.ai/schemas/mistral/fine-tuning-job.json", "title": "Mistral Fine-Tuning Job", "description": "A fine-tuning job for customizing a Mistral model on user-provided training data, with configurable hyperparameters, status tracking, and integration support.", "type": "object", "required": ["id", "model", "status"], "properties": { "id": { "type": "string", "format": "uuid", "description": "Unique identifier for the fine-tuning job" }, "auto_start": { "type": "boolean", "description": "Whether training starts automatically after validation" }, "model": { "type": "string", "description": "The base model being fine-tuned" }, "status": { "type": "string", "enum": [ "QUEUED", "STARTED", "VALIDATING", "VALIDATED", "RUNNING", "FAILED_VALIDATION", "FAILED", "SUCCESS", "CANCELLED", "CANCELLATION_REQUESTED" ], "description": "Current status of the fine-tuning job" }, "job_type": { "type": "string", "description": "Type of fine-tuning job (e.g., text, vision, classifier)" }, "created_at": { "type": "integer", "description": "Unix timestamp when the job was created" }, "modified_at": { "type": "integer", "description": "Unix timestamp when the job was last modified" }, "training_files": { "type": "array", "items": { "type": "string", "format": "uuid" }, "description": "IDs of training data files" }, "validation_files": { "type": "array", "items": { "type": "string", "format": "uuid" }, "description": "IDs of validation data files" }, "hyperparameters": { "$ref": "#/$defs/Hyperparameters" }, "fine_tuned_model": { "type": ["string", "null"], "description": "ID of the resulting fine-tuned model upon completion" }, "suffix": { "type": "string", "maxLength": 18, "description": "User-specified suffix appended to the fine-tuned model name" }, "integrations": { "type": "array", "items": { "$ref": "#/$defs/Integration" }, "description": "Third-party integrations for experiment tracking" } }, "$defs": { "Hyperparameters": { "type": "object", "description": "Hyperparameters for the fine-tuning job", "properties": { "training_steps": { "type": "integer", "minimum": 1, "description": "Number of training steps" }, "learning_rate": { "type": "number", "minimum": 0, "description": "Learning rate" }, "weight_decay": { "type": "number", "minimum": 0, "description": "Weight decay regularization factor" }, "warmup_fraction": { "type": "number", "minimum": 0, "maximum": 1, "description": "Fraction of total steps for learning rate warmup" }, "epochs": { "type": "number", "minimum": 0, "description": "Number of training epochs" }, "seq_len": { "type": "integer", "description": "Maximum sequence length for training" } } }, "Integration": { "type": "object", "description": "A third-party integration for experiment tracking", "required": ["type"], "properties": { "type": { "type": "string", "enum": ["wandb"], "description": "Integration provider type" }, "project": { "type": "string", "description": "Project name in the integration platform" }, "name": { "type": "string", "description": "Run name in the integration platform" } } } } }