https://pypi.org/project/guarded-hotshard/ Recent updates to the Python Package Index for guarded-hotshard en https://pypi.org/project/guarded-hotshard/0.2.0/ Tenant-aware request scheduling for LLM inference. Cuts premium-tenant p99 latency 60-70% on a contested GPU at <5% extra cost. Sat, 02 May 2026 05:28:37 GMT https://pypi.org/project/guarded-hotshard/0.1.0/ Tenant-aware request scheduling for LLM inference. Cuts premium-tenant p99 latency 60-70% on a contested GPU at <5% extra cost. Sat, 02 May 2026 05:12:57 GMT