Lifecycle

A Fine-tuning Job progresses through several statuses:

• pending: The workload has been registered and its creation process has begun.
• scheduling: FlexAI is searching for the suitable hardware resources to run the workload.
• building: The Runtime environment is being prepared, including cloning the source repository on the specified revision and installing the project's dependencies.
• in progress: The workload is currently executing the code defined in the entry point file.
• enqueued: The workload is waiting for resources to become available.
• succeeded: The workload completed successfully and output artifacts are ready for download.
• failed: The workload did not complete successfully. Common reasons include missing files, dependency errors, or exceeding time limits.
• stopped: The workload was stopped by the user before completion.

Terminal statuses (succeeded, failed, stopped) mean the job is finished and can be deleted if desired.