> ## 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. # Get training job queue context > Returns the (org, gpu_type) capacity pool the job was gated by, jobs that were holding GPU capacity in that pool when this job was submitted, and every TrainingJobStatus event in [submitted_at, released_at] for those jobs — useful for understanding why a job sat in PENDING. Caller must be an org admin and the job must belong to the caller's org. ## OpenAPI ````yaml get /v1/training/jobs/{training_job_id}/queue_context 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/training/jobs/{training_job_id}/queue_context: parameters: - $ref: '#/components/parameters/training_job_id' get: summary: Reconstruct queue context for a training job. description: >- Returns the (org, gpu_type) capacity pool the job was gated by, jobs that were holding GPU capacity in that pool when this job was submitted, and every TrainingJobStatus event in [submitted_at, released_at] for those jobs — useful for understanding why a job sat in PENDING. Caller must be an org admin and the job must belong to the caller's org. responses: '200': description: '' content: application/json: schema: $ref: '#/components/schemas/GetTrainingJobQueueContextResponseV1' x-codeSamples: - lang: bash source: > curl --request GET \ --url https://api.baseten.co/v1/training/jobs/{training_job_id}/queue_context \ --header "Authorization: Bearer $BASETEN_API_KEY" - lang: python source: >- import requests import os API_KEY = os.environ.get("BASETEN_API_KEY", "") url = "https://api.baseten.co/v1/training/jobs/{training_job_id}/queue_context" headers = {"Authorization": f"Bearer {API_KEY}"} response = requests.request( "GET", url, headers=headers, json={} ) print(response.text) components: parameters: training_job_id: schema: type: string name: training_job_id in: path required: true schemas: GetTrainingJobQueueContextResponseV1: description: >- Read-only diagnostic for a training job's PENDING window. Returns the (org, gpu_type) capacity pool the job was gated by, jobs that were holding GPU capacity in that pool when this job was submitted, and every status event in [submitted_at, released_at] for those jobs (or up to "now" if the target is still PENDING). properties: target_job_id: description: Hashid of the target training job title: Target Job Id type: string target_job_name: anyOf: - type: string - type: 'null' default: null description: Target job's name title: Target Job Name gpu_type: description: GPU type the target requested title: Gpu Type type: string requested_gpus: description: GPUs the target requested (gpu_count * effective_node_count) title: Requested Gpus type: integer submitted_at: description: When the job row was inserted (= API POST time) format: date-time title: Submitted At type: string released_at: anyOf: - format: date-time type: string - type: 'null' default: null description: >- When the job's TRAINING_JOB_CREATED status was set, i.e. the moment it was released from PENDING. None if still PENDING. title: Released At pending_seconds: anyOf: - type: integer - type: 'null' default: null description: released_at - submitted_at in seconds. None if still PENDING. title: Pending Seconds org_capacity: anyOf: - $ref: '#/components/schemas/CapacityAtSubmitV1' - type: 'null' default: null description: Org-level cap for (org, gpu_type). None if no cap is configured. team_capacity: anyOf: - $ref: '#/components/schemas/CapacityAtSubmitV1' - type: 'null' default: null description: >- Team-level cap for (team, gpu_type). None if no team cap is configured. active_at_submit: description: >- Jobs in the same (org, gpu_type) pool that were holding capacity at submitted_at items: $ref: '#/components/schemas/ActiveJobAtSubmitV1' title: Active At Submit type: array pending_ahead_at_submit: description: >- PENDING jobs in the same (org, gpu_type) pool that were ahead of the target in dequeue FIFO order at submitted_at (priority DESC then created ASC). These also block the target's release. items: $ref: '#/components/schemas/PendingJobAheadAtSubmitV1' title: Pending Ahead At Submit type: array events: description: >- Every status event in [submitted_at, events_window_end] for the target job, every job in active_at_submit, and every job in pending_ahead_at_submit, oldest first. items: $ref: '#/components/schemas/QueueEventV1' title: Events type: array events_window_end: description: released_at if set, else 'now' (events ongoing) format: date-time title: Events Window End type: string required: - target_job_id - gpu_type - requested_gpus - submitted_at - active_at_submit - pending_ahead_at_submit - events - events_window_end title: GetTrainingJobQueueContextResponseV1 type: object CapacityAtSubmitV1: description: >- A GPU capacity row as it stands now, with ``last_modified`` so callers can judge whether the value matches what the dequeue gate saw at submit time. Capacity rows are not historicized: edits overwrite in place. Compare ``last_modified`` against the response's ``submitted_at`` — if it's later, the value may have changed. properties: gpu_type: description: GPU type identifier (e.g. H100, A100-40GB) title: Gpu Type type: string max_gpus: description: Current max concurrent GPUs of this type title: Max Gpus type: integer min_gpus: anyOf: - type: integer - type: 'null' default: null description: Current baseline GPU allocation, if configured title: Min Gpus last_modified: description: When the capacity row was last modified format: date-time title: Last Modified type: string required: - gpu_type - max_gpus - last_modified title: CapacityAtSubmitV1 type: object ActiveJobAtSubmitV1: description: |- One other job in the same (org, gpu_type) pool that was holding GPU capacity at the moment the target was submitted. properties: training_job_id: description: Hashid of the other training job title: Training Job Id type: string training_job_name: anyOf: - type: string - type: 'null' default: null description: Other job's name title: Training Job Name instance_type_name: description: Instance type of the other job title: Instance Type Name type: string total_gpus: description: gpu_count * effective_node_count title: Total Gpus type: integer workload_plane_name: description: Workload plane the other job was on title: Workload Plane Name type: string status_at_submit: description: Other job's status as of submitted_at (one of ACTIVE_STATES) title: Status At Submit type: string status_set_at: description: When that status was set format: date-time title: Status Set At type: string required: - training_job_id - instance_type_name - total_gpus - workload_plane_name - status_at_submit - status_set_at title: ActiveJobAtSubmitV1 type: object PendingJobAheadAtSubmitV1: description: |- A PENDING job in the same (org, gpu_type) pool that was ahead of the target in dequeue FIFO order at submitted_at — higher priority, or same priority and earlier submission. properties: training_job_id: description: Hashid of the other training job title: Training Job Id type: string training_job_name: anyOf: - type: string - type: 'null' default: null description: Other job's name title: Training Job Name instance_type_name: description: Instance type of the other job title: Instance Type Name type: string requested_gpus: description: gpu_count * effective_node_count title: Requested Gpus type: integer priority: description: Effective priority (NULL coalesced to 0) title: Priority type: integer submitted_at: description: The other job's submission time format: date-time title: Submitted At type: string required: - training_job_id - instance_type_name - requested_gpus - priority - submitted_at title: PendingJobAheadAtSubmitV1 type: object QueueEventV1: description: A single ``TrainingJobStatus`` row inside the queue-context window. properties: training_job_id: description: Hashid of the training job this event is for title: Training Job Id type: string training_job_name: anyOf: - type: string - type: 'null' default: null description: Job name title: Training Job Name status: description: TrainingJobStatus.Name value title: Status type: string created: description: When the status row was inserted format: date-time title: Created type: string event_message: anyOf: - type: string - type: 'null' default: null description: Human-readable event message from the status metadata title: Event Message exit_code: anyOf: - type: integer - type: 'null' default: null description: Exit code from the status metadata, if any title: Exit Code required: - training_job_id - status - created title: QueueEventV1 type: object securitySchemes: BearerAuth: type: http scheme: bearer description: >- Send `Authorization: Bearer `. The legacy `Authorization: Api-Key ` scheme is also accepted. ````