Hi everyone!
To fully leverage the new Devshards scaling and begin processing a higher volume of inference requests, a minimal update to your proxy configuration is required.
Please complete the following steps:
- Update docker-compose.yml
Add subnet to GONKA_API_EXEMPT_ROUTES values:
… - GONKA_API_EXEMPT_ROUTES=chat inference poc/proofs subnet ….
Full diff: Comparing main...proxy-shards · gonka-ai/gonka · GitHub
- Restart proxy container
shell
source config.env && docker compose up proxy -d --force-recreate
Please also ensure that there are no other strict rate limits applied to the /v1/subnet/* path in your infrastructure. Thanks!