> ## Documentation Index > Fetch the complete documentation index at: https://docs.together.ai/llms.txt > Use this file to discover all available pages before exploring further. # Extend model resources expiry > Keeps a ready model resource from stopping for inactivity for at least the given number of seconds from now. It moves the resource's `expires_at` to at least that time, and never earlier. Every operation on the resource's sessions already keeps it up for `inactive_timeout` seconds, so use this only to stay up longer without operations, for example while you compute rewards. Fails if the resource is not ready. ## OpenAPI ````yaml openapi.yaml POST /v1/rl/model-resources/{model_resources_id}/extend-expiry openapi: 3.1.0 info: title: Together APIs description: The Together REST API. See https://docs.together.ai for more details. version: 2.0.0 termsOfService: https://www.together.ai/terms-of-service contact: name: Together Support url: https://www.together.ai/contact license: name: MIT url: https://github.com/togethercomputer/openapi/blob/main/LICENSE servers: - url: https://api.together.ai description: Default environment for APIs security: - bearerAuth: [] paths: /v1/rl/model-resources/{model_resources_id}/extend-expiry: post: tags: - RL summary: Extend model resources expiry description: >- Keeps a ready model resource from stopping for inactivity for at least the given number of seconds from now. It moves the resource's `expires_at` to at least that time, and never earlier. Every operation on the resource's sessions already keeps it up for `inactive_timeout` seconds, so use this only to stay up longer without operations, for example while you compute rewards. Fails if the resource is not ready. operationId: extendModelResourcesExpiry parameters: - name: model_resources_id in: path required: true schema: description: ID of the model resource type: string default: false requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/RL.ExtendModelResourcesExpiryRequest' responses: '200': description: Model resource details content: application/json: schema: $ref: '#/components/schemas/RL.ModelResources' default: description: An unexpected error response. content: application/json: schema: $ref: '#/components/schemas/ErrorData' x-codeSamples: - lang: Python label: Together AI SDK (Python) source: | from together import Together client = Together() resource = client.post_training.model_resources.extend_expiry( "", seconds=7200, ) print(resource.status, resource.expires_at) components: schemas: RL.ExtendModelResourcesExpiryRequest: type: object description: Request to keep a model resource from stopping for inactivity required: - seconds properties: seconds: type: integer minimum: 60 maximum: 86400 example: 7200 description: Seconds from now the resource must not stop for inactivity. RL.ModelResources: type: object description: Allocated GPU resources that training sessions attach to required: - id - status - status_details - base_model - compute_config - lora_enabled - created_at - updated_at - optimizer_config - created_by - inactive_timeout properties: id: type: string example: 123e4567-e89b-12d3-a456-426614174000 description: Unique identifier for the model resource status: $ref: '#/components/schemas/RL.ModelResourcesStatus' example: MODEL_RESOURCES_STATUS_READY description: Lifecycle status of the model resource error: $ref: '#/components/schemas/RL.ModelResourcesError' description: >- Structured detail for the model resource's current error. Set when the resource is in an error state. status_details: $ref: '#/components/schemas/RL.ModelResourcesStatusDetails' description: >- Details of the resource's current status, such as its place in the queue. base_model: type: string example: Qwen/Qwen3.5-4B description: Base model the resource is provisioned for base_weights_ref: type: string example: together://ml_A1b2C3d4E5f6G7h8I9j0K@rv_A1b2C3d4E5f6G7h8I9j0K description: >- Model-registry model whose weights the resource trains in place of the base model's own, pinned to the revision resolved when the resource was created. Absent when the resource trains the base model's own weights. compute_config: $ref: '#/components/schemas/RL.ComputeConfig' description: Compute layout provisioned for the resource. lora_enabled: type: boolean example: true description: Whether the resource hosts LoRA sessions or a full-weight session created_at: type: string format: date-time example: '2026-01-02T00:00:00Z' description: Timestamp when the model resource was created updated_at: type: string format: date-time example: '2026-01-02T00:00:05Z' description: Timestamp when the model resource was last updated optimizer_config: $ref: '#/components/schemas/RL.OptimizerConfig' description: Optimizer configuration for this resource. created_by: type: string example: user_123 description: ID of the user who created the model resource inactive_timeout: type: integer example: 3600 description: >- Seconds without activity after which the resource stops and its sessions expire. Every operation on any of its sessions restarts the count. expires_at: type: string format: date-time example: '2026-01-02T01:00:05Z' description: >- The earliest time the resource can stop for inactivity, which also expires its sessions. It is the resource's last activity plus `inactive_timeout` seconds. Becoming ready, a session start and every operation on any of its sessions count as activity, and the resource does not stop while an operation is running. An extend expiry call can move it later. It never moves earlier. Refreshed about once a minute. Set while the resource is ready. ErrorData: type: object required: - error properties: error: type: object properties: message: type: string nullable: false type: type: string nullable: false param: type: string nullable: true default: null code: type: string nullable: true default: null required: - type - message RL.ModelResourcesStatus: description: Lifecycle status of a model resource type: string enum: - MODEL_RESOURCES_STATUS_PENDING - MODEL_RESOURCES_STATUS_CREATING - MODEL_RESOURCES_STATUS_READY - MODEL_RESOURCES_STATUS_ERROR - MODEL_RESOURCES_STATUS_STOPPED - MODEL_RESOURCES_STATUS_STOPPING RL.ModelResourcesError: type: object description: Structured detail for the model resource's current error required: - code - message - occurred_at properties: code: $ref: '#/components/schemas/RL.ModelResourcesErrorCode' example: MODEL_RESOURCES_ERROR_CODE_CAPACITY_WAIT_TIMEOUT description: Finite machine-readable reason code for UI branching message: type: string example: Timed out waiting for cluster capacity. Try again later. description: User-safe human-readable detail for the current status occurred_at: type: string format: date-time example: '2026-01-02T00:00:05Z' description: Timestamp when this error was reported RL.ModelResourcesStatusDetails: type: object description: Details of a model resource's current status properties: queue_position: type: integer format: int64 example: 3 description: >- Estimated place of the resource in the queue: the number of older resources of the same GPU type still waiting for GPUs, plus one. Set only while the resource waits for GPUs. RL.ComputeConfig: type: object description: Compute layout for a model resource required: - num_generator_replicas properties: num_generator_replicas: type: integer format: int32 example: 1 description: >- Number of generator replicas. 0 means the resource runs the trainer only, with no generator. gpu_type: type: string example: B200-SXM enum: - H100-80GB - B200-SXM description: GPU type selected for this resource. RL.OptimizerConfig: type: object description: Optimizer configuration properties: adam: $ref: '#/components/schemas/RL.AdamConfig' description: Use the Adam optimizer. muon: $ref: '#/components/schemas/RL.MuonConfig' description: Use the Muon optimizer. RL.ModelResourcesErrorCode: type: string description: Finite machine-readable model resource lifecycle error code enum: - MODEL_RESOURCES_ERROR_CODE_CAPACITY_WAIT_TIMEOUT - MODEL_RESOURCES_ERROR_CODE_PROVISIONING_FAILED - MODEL_RESOURCES_ERROR_CODE_RESOURCE_UNAVAILABLE RL.AdamConfig: type: object description: Adam optimizer configuration properties: {} additionalProperties: false x-stainless-empty-object: true RL.MuonConfig: type: object description: Advanced configuration for the Muon optimizer. properties: scaling_strategy: $ref: '#/components/schemas/RL.MuonScalingStrategy' example: MUON_SCALING_STRATEGY_ORIGINAL default: MUON_SCALING_STRATEGY_MATCH_ADAM description: Scaling strategy for the Muon optimizer. RL.MuonScalingStrategy: type: string enum: - MUON_SCALING_STRATEGY_UNSPECIFIED - MUON_SCALING_STRATEGY_MATCH_ADAM - MUON_SCALING_STRATEGY_ORIGINAL default: MUON_SCALING_STRATEGY_UNSPECIFIED securitySchemes: bearerAuth: type: http scheme: bearer x-bearer-format: bearer x-default: default ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.