> ## Documentation Index > Fetch the complete documentation index at: https://docs.baseten.co/llms.txt > Use this file to discover all available pages before exploring further. # Create a run > Creates a Loops run with an associated sampler in the given session. ## OpenAPI ````yaml post /v1/loops/runs openapi: 3.1.0 info: description: REST API for management of Baseten resources title: Baseten management API version: 1.0.0 servers: - url: https://api.baseten.co security: - BearerAuth: [] paths: /v1/loops/runs: post: summary: Creates a Loops run description: Creates a Loops run with an associated sampler in the given session. requestBody: content: application/json: schema: $ref: '#/components/schemas/CreateLoopsRunRequestV1' required: true responses: '200': description: '' content: application/json: schema: $ref: '#/components/schemas/CreateLoopsRunResponseV1' x-codeSamples: - lang: bash source: |- curl --request POST \ --url https://api.baseten.co/v1/loops/runs \ --header "Authorization: Bearer $BASETEN_API_KEY" \ --data '{ "session_id": null, "base_model": null, "name": null, "max_seq_len": null, "seed": null, "availability_model": "spot", "path": "bt://loops:k4q95w5/weights/step-100", "reuse_from_run_id": null, "reuse_from_session_id": null }' - lang: python source: |- import requests import os API_KEY = os.environ.get("BASETEN_API_KEY", "") url = "https://api.baseten.co/v1/loops/runs" headers = {"Authorization": f"Bearer {API_KEY}"} response = requests.request( "POST", url, headers=headers, json={'session_id': None, 'base_model': None, 'name': None, 'max_seq_len': None, 'seed': None, 'availability_model': 'spot', 'path': 'bt://loops:k4q95w5/weights/step-100', 'reuse_from_run_id': None, 'reuse_from_session_id': None} ) print(response.text) components: schemas: CreateLoopsRunRequestV1: properties: session_id: description: ID of the Loops session this run belongs to. title: Session Id type: string base_model: description: Base model ID (e.g. 'Qwen/Qwen3-8B'). title: Base Model type: string name: anyOf: - maxLength: 255 minLength: 1 type: string - type: 'null' default: null description: >- Optional display name for the run. Defaults to the base model name when omitted. title: Name max_seq_len: anyOf: - type: integer - type: 'null' default: null description: >- Maximum prompt length (in tokens) the run must handle. Set this to the longest training example you plan to send. Defaults to the maximum supported by the model configuration. title: Max Seq Len lora_rank: default: 64 description: LoRA rank. minimum: 1 title: Lora Rank type: integer seed: anyOf: - type: integer - type: 'null' default: null description: Random seed for reproducibility. title: Seed scale_down_delay_seconds: default: 900 description: >- Seconds of inactivity before the run scales to zero. Must be between 1 and 3600 (1 hour). Defaults to 900 (15 minutes). exclusiveMinimum: 0 maximum: 3600 title: Scale Down Delay Seconds type: integer availability_model: $ref: '#/components/schemas/V1AvailabilityModel' default: dedicated description: >- Capacity the trainer runs on. 'dedicated' is not preempted. 'spot' runs below inference and reaches idle reserved capacity, but the run is stopped if its GPUs are reclaimed and cannot be resumed. examples: - spot replicas: default: 1 description: >- Number of data-parallel trainer replicas. Each replica is one full copy of the model's preset node group, so the trainer deployment runs (preset node_count * replicas) nodes (e.g. replicas=4 on a 4-node preset → 16 nodes, 4 DP workers). Must be a positive integer. Defaults to 1. minimum: 1 title: Replicas type: integer path: anyOf: - type: string - type: 'null' default: null description: >- Optional bt:// URI of an existing checkpoint to load weights from on startup. Form: bt://loops:/weights/. examples: - bt://loops:k4q95w5/weights/step-100 title: Path reuse_from_run_id: anyOf: - type: string - type: 'null' default: null description: >- Optional ID of a prior Loops run whose trainer and/or sampler should be reused for this run instead of provisioning fresh. The prior run must use the same base model and belong to the same team. title: Reuse From Run Id reuse_from_session_id: anyOf: - type: string - type: 'null' default: null description: >- Optional ID of a prior Loops session whose trainer and/or sampler should be reused for this run. Deprecated in favor of reuse_from_run_id. title: Reuse From Session Id required: - session_id - base_model title: CreateLoopsRunRequestV1 type: object CreateLoopsRunResponseV1: properties: run: $ref: '#/components/schemas/LoopsRunV1' required: - run title: CreateLoopsRunResponseV1 type: object V1AvailabilityModel: description: >- Capacity guarantee under which a training job is scheduled. ``DEDICATED`` is on-demand capacity that is not preempted (the default). ``SPOT`` is interruptible capacity that may be preempted; the user is responsible for checkpointing their own progress. A managed/resumable model where the platform handles checkpoint/resume on its own is intentionally not defined yet; it is planned for a future milestone. enum: - dedicated - spot title: V1AvailabilityModel type: string LoopsRunV1: properties: id: description: The run ID. title: Id type: string session_id: description: The session ID this run belongs to. title: Session Id type: string deployment_id: anyOf: - type: string - type: 'null' default: null description: The ID of the Loops deployment the run executes on, if it has one. title: Deployment Id name: description: The run's display name. title: Name type: string base_model: description: The HuggingFace base model the run is fine-tuning. title: Base Model type: string base_url: description: The run's base URL. title: Base Url type: string created_at: description: Time the run was created in ISO 8601 format format: date-time title: Created At type: string status: $ref: '#/components/schemas/LoopsRunStatusV1' description: The run's current status. user: $ref: '#/components/schemas/UserV1' description: The user who owns the run. sampler: anyOf: - $ref: '#/components/schemas/LoopsSamplerV1' - type: 'null' default: null description: >- The sampler bound to this run, or null for a trainer-only run that has not yet created a sampler. required: - id - session_id - name - base_model - base_url - created_at - status - user title: LoopsRunV1 type: object LoopsRunStatusV1: description: The current status of a Loops run. properties: name: $ref: '#/components/schemas/LoopsRunStatusNameV1' description: >- ACTIVE while the run is live; INACTIVE once replaced by a newer run or shut down. required: - name title: LoopsRunStatusV1 type: object UserV1: description: A user. properties: email: anyOf: - type: string - type: 'null' default: null description: Email of the user. title: Email title: UserV1 type: object LoopsSamplerV1: properties: id: title: Id type: string base_url: title: Base Url type: string base_model: description: The HuggingFace base model the sampler is serving. title: Base Model type: string created_at: description: Time the sampler was created in ISO 8601 format format: date-time title: Created At type: string model_id: description: Hashid of the underlying Baseten model. title: Model Id type: string deployment_id: description: Hashid of the specific model deployment (version). title: Deployment Id type: string status: $ref: '#/components/schemas/LoopsSamplerStatusV1' description: The sampler's current status. user: $ref: '#/components/schemas/UserV1' description: The user who owns the sampler. instance_type: anyOf: - $ref: '#/components/schemas/InstanceTypeV1' - type: 'null' default: null description: Instance type serving the sampler. node_count: default: 1 description: Number of nodes serving the sampler. title: Node Count type: integer required: - id - base_url - base_model - created_at - model_id - deployment_id - status - user title: LoopsSamplerV1 type: object LoopsRunStatusNameV1: description: 'A Loops run''s lifecycle state: ACTIVE or INACTIVE.' enum: - ACTIVE - INACTIVE title: LoopsRunStatusNameV1 type: string LoopsSamplerStatusV1: description: The current status of a Loops sampler. properties: name: $ref: '#/components/schemas/DeploymentStatusV1' description: The current status of the Loops sampler. required: - name title: LoopsSamplerStatusV1 type: object InstanceTypeV1: description: An instance type. properties: id: description: Identifier string for the instance type title: Id type: string name: description: Display name of the instance type title: Name type: string memory_limit_mib: description: Memory limit of the instance type in Mebibytes title: Memory Limit Mib type: integer millicpu_limit: description: CPU limit of the instance type in millicpu title: Millicpu Limit type: integer gpu_count: description: Number of GPUs on the instance type title: Gpu Count type: integer gpu_type: anyOf: - type: string - type: 'null' description: Type of GPU on the instance type title: Gpu Type gpu_memory_limit_mib: anyOf: - type: integer - type: 'null' description: Memory limit of the GPU on the instance type in Mebibytes title: Gpu Memory Limit Mib required: - id - name - memory_limit_mib - millicpu_limit - gpu_count - gpu_type - gpu_memory_limit_mib title: InstanceTypeV1 type: object DeploymentStatusV1: description: The status of a deployment. enum: - BUILDING - DEPLOYING - DEPLOY_FAILED - LOADING_MODEL - ACTIVE - UNHEALTHY - BUILD_FAILED - BUILD_STOPPED - DEACTIVATING - INACTIVE - FAILED - UPDATING - SCALED_TO_ZERO - WAKING_UP title: DeploymentStatusV1 type: string securitySchemes: BearerAuth: type: http scheme: bearer description: >- Send `Authorization: Bearer `. The legacy `Authorization: Api-Key ` scheme is also accepted. ````