Configure API keys, training defaults, and storage.
Used to push models to the Hub and launch training jobs.
Used to call Grok models for dataset generation and evaluation.
Keys are stored server-side only
Pre-selected hardware tier when creating a new pipeline.
Pre-selected training method for new pipelines.
Automatically enable Hub push when creating new pipelines.
Cloudflare R2 bucket used for dataset and model artifact storage.
Read-only. Set via CLOUDFLARE_R2_ENDPOINT environment variable.
Storage Usage
—
Datasets
—
Model Artifacts
—
Total Used
Storage stats will be available once R2 integration is configured.