Rate limits ↗
noOriginal Documentation
Documentation Index#
Fetch the complete documentation index at: https://docs.wandb.ai/llms.txt Use this file to discover all available pages before exploring further.
Optional rate limits on Self-Managed instances for stability
You can configure rate limits on your W&B Self-Managed instance to maintain stability and prevent one user or workload from negatively impacting others. These limits are optional. If you do not set them, the instance does not enforce the limits described below.
Default limits and notification#
The following default limits help maintain platform stability:
| Limit | Default | Scope |
|---|---|---|
| Filestream requests per second | 500 | Project |
| Filestream ingestion per second | 120 MB | Project |
| Filestream requests per second | 2 | Run |
| Run creation requests per second | 80 | Project |
When a limit is exceeded, the W&B SDK returns HTTP response 429, and the message HTTP 429: rate limited exceeded appears in the SDK logs.
Filesystem rate limits never cause logging to crash or fail. When the SDK receives a
429response on a filestream request, it will back off and retry the rate-limited request as-is, while subsequent updates accumulate.Run creation rate limits block further training.
HTTP 429: rate limit exceededThese defaults are recommended for typical production workloads. If your workloads consistently exceed these limits, you can configure these limits (or leave them unset) according to your instance size and shared usage. For configuration details, refer to Deploy W&B with Kubernetes Operator or contact W&B support.