We have a usage tier 5 organization with token limits of 10 Million TPM and request limit of 10k RPM for text-embedding-ada-002. Our usage dashboard shows total usage for today to be ~6300 requests and 5.6M tokens for text-embedding-ada-002. Despite our usage being well under the limit, we are being rate limited for the Embeddings API for text-embedding-ada-002.
Am I missing something here or is this some unusual behavior?