mirror of https://github.com/langchain-ai/langchain-pinecone.git synced 2026-07-19 13:54:50 -04:00

T

Jen Hamon a9fedf08ab fix(embeddings): use query_params in aembed_query

The async `aembed_query` was passing `self.document_params` (input_type:
"passage") to `_aembed_texts` while the sync `embed_query` correctly passes
`self.query_params` (input_type: "query"). This caused async query embeddings
to be embedded as passages, which degrades retrieval quality for async callers.

Change `parameters=self.document_params` to `parameters=self.query_params` in
`PineconeEmbeddings.aembed_query` (one line). The sparse twin
`PineconeSparseEmbeddings.aembed_query` already used `query_params` correctly.

Add three regression tests in `TestQueryParamsParity`:
- `test_embed_query_uses_query_params` — pins the sync contract
- `test_aembed_query_uses_query_params` — fails before the fix, passes after
- `test_sparse_aembed_query_uses_query_params` — pins the sparse async contract

2026-06-12 19:41:38 +00:00

.github

chore(deps): bump requests 2.32.4 to 2.33.0 + fix CI for Dependabot (#104 )

2026-04-03 21:03:42 -07:00

examples

feat: new examples and v0.2.13 release (#92 )

2025-11-02 20:01:07 +01:00

libs/pinecone

fix(embeddings): use query_params in aembed_query

2026-06-12 19:41:38 +00:00

.gitignore

Added dynamic embedding and reranker models (#65 ) (#66 )