Context
I have a large Obsidian vault (~18,500 markdown files, ~1.8 GB) and upgraded from enzyme v0.25.3 to v0.4.6.
With v0.25.3, all documents were chunked and embedded (416,379 chunks). With v0.4.6, only ~1,000-3,000 "recent" documents get embeddings. The rest are indexed (entities, tags, links) but not embedded.
Problem
For my use case (research vault with academic papers, project notes, clippings spanning 10+ years), I need all documents embedded for semantic search via enzyme catalyze. Many important connections are between older papers and current projects. Limiting to recent docs misses these cross-temporal patterns.
Request
Could you add a config option or flag to embed all documents? Something like:
# enzyme-config.yaml
embedding:
scope: all # or "recent" (default)
Or a CLI flag:
Current workaround
Running v0.25.3 in parallel for full embeddings, but the DB format is incompatible with v0.4.6.
Thanks for the great tool!
Context
I have a large Obsidian vault (~18,500 markdown files, ~1.8 GB) and upgraded from enzyme v0.25.3 to v0.4.6.
With v0.25.3, all documents were chunked and embedded (416,379 chunks). With v0.4.6, only ~1,000-3,000 "recent" documents get embeddings. The rest are indexed (entities, tags, links) but not embedded.
Problem
For my use case (research vault with academic papers, project notes, clippings spanning 10+ years), I need all documents embedded for semantic search via
enzyme catalyze. Many important connections are between older papers and current projects. Limiting to recent docs misses these cross-temporal patterns.Request
Could you add a config option or flag to embed all documents? Something like:
Or a CLI flag:
Current workaround
Running v0.25.3 in parallel for full embeddings, but the DB format is incompatible with v0.4.6.
Thanks for the great tool!