# CLaaS

## Docs

- [Configuration Reference](https://docs.openclaas.com/configuration.md): Complete Hydra configs and environment variables for all CLaaS backends and eval
- [Dashboards](https://docs.openclaas.com/dashboards.md): Built-in web dashboards for training metrics and eval results
- [Metrics](https://docs.openclaas.com/eval/metrics.md): Eval metrics, collapse detection, and compliance verifiers
- [Eval Harness Overview](https://docs.openclaas.com/eval/overview.md): Automated evaluation for SDPO continual learning
- [Configuration & Running](https://docs.openclaas.com/eval/running.md): Configure and run the CLaaS eval harness
- [CLaaS Documentation](https://docs.openclaas.com/index.md): Continual Learning as a Service. Real-time model personalization via self-distillation
- [Quick Start](https://docs.openclaas.com/quickstart.md): Get CLaaS running with Local, Tinker, or Modal backends
- [Local GPU Backend](https://docs.openclaas.com/training/local.md): Run CLaaS with local GPU training and vLLM inference
- [Modal Cloud Backend](https://docs.openclaas.com/training/modal.md): Deploy CLaaS to Modal for remote GPU training on L40S
- [Training Backends Overview](https://docs.openclaas.com/training/overview.md): Compare Local, Tinker, and Modal training backends
- [Tinker SDK Backend](https://docs.openclaas.com/training/tinker.md): Run CLaaS with Tinker's hosted inference and training - no GPU required

## OpenAPI Specs

- [openapi](https://docs.openclaas.com/api-reference/openapi.json)