langchain-ai/langchain-google

First Batch of results in `VertexAIEmbeddings` does not respect the dimensions passed for embedding

Open

#1,089 opened on Aug 22, 2025

View on GitHubContributor guide
 (2 comments) (0 reactions) (0 assignees)Python (461 forks)auto 404
help wantedinvestigatevertexai

Repository metrics

Stars
 (392 stars)
PR merge metrics
 (PR metrics pending)

Description

https://github.com/langchain-ai/langchain-google/blob/a2a05f568ddc5a823617e70d5b379187938893f4/libs/vertexai/langchain_google_vertexai/embeddings.py#L367

In here, it seems that the dimensions is not passed for the first batch to be processed based on the embedding dimensions passed by the client. This Creates a bug when custom dimensions needs to be embedded from VertexAI Embeddings (eg. 768 dimension embedding from gemini-embedding-001)

The fix is simple. pass in the dimensions forward:

first_result = self._get_embeddings_with_retry(
                        first_batch, embeddings_type, dimensions
                    )

Contributor guide

Research direction
Pass the `dimensions` parameter to the ` get embeddings with retry` call for the first batch in the `embed documents` method, matching the second batch call.
Tech stack
python
Domain
aibackend
Issue type
Bug
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
1
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
Under 1 hour
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
None
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
90