Expand description
Persistent compilation cache for TensorLogic
This module provides disk-based caching of compiled graphs to speed up repeated compilations. The cache uses LRU (Least Recently Used) eviction policy with compression support.
Structsยง
- Cache
Analytics - Cache analytics with derived metrics and recommendations
- Cache
Entry - Cache entry containing the compiled graph and metadata
- Cache
Stats - Cache statistics with LRU metrics
- Cache
Warmup Result - Result of a cache warmup operation
- Compilation
Cache - Persistent compilation cache with LRU eviction