Repository navigation
Conversation
…econe `xerj autoindex qdrant://host:port/collection`, `chroma://...`, `pinecone://index` and a pgvector table read the source's vectors, ids and payload, create a dense_vector mapping with the source's dimensions and similarity, and write through _bulk with 4 parallel writers. A journal makes the import resumable; progress shows an ETA; a dry run takes a row limit. After the import the importer compares its count with the source and warns on any difference. Live checks against real containers (same host): Qdrant 20,000 points, Chroma 3,000 points, pgvector 5,000 rows; each sample of 100 vectors returned the same top-1 from XERJ kNN as from the source (100/100). Pinecone: unit tests against a mocked HTTP server (no account). The pgvector live test found a paging bug (4,022 of 5,000 rows imported); fixed, and the count-mismatch warning was added so a short import is never silent. Not done: the 1M x 384-d throughput and memory measurement; MySQL is not a vector source. Tests: cargo test -p xerj-autoindex vecimport, 12 passed, 4 live-container tests ignored without Docker.
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Imports vectors, ids and payload from Qdrant, Chroma, pgvector and Pinecone into XERJ with an auto-created dense_vector mapping, resumable journal, progress/ETA and a post-import count check.
Live checks against real containers: Qdrant 20,000 points, Chroma 3,000, pgvector 5,000 rows; kNN top-1 agreement with the source 100/100 on a 100-vector sample each. Pinecone: mocked HTTP server only (no account). Found and fixed a pgvector paging bug (4,022 of 5,000 imported before the fix).
Not done: the 1M-vector throughput/memory measurement.
🤖 Generated with Claude Code