github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 08:02:14 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #11213] Add support for custom vector dimensions in embedding #69445

New Issue
Closed
opened 2026-05-04 18:07:31 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-05-04 18:07:31 -05:00
Owner
Copy Link

Originally created by @bgeneto on GitHub (Jun 26, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/11213

New embedding models like Qwen3 Embedding (e.g. Qwen3-Embedding-8B) has "MRL Support", meaning one can configure the vector dimension for the final embedding. The 8B model, for example, has a default dimension of 4096. This, unfortunately, implies that it cannot be used with pgvector from Postgres, unless Ollama adds a new dimensions parameter (like DeepInfra) for supporting changing the default model dimension. An example usage of this parameter is here.

Originally created by @bgeneto on GitHub (Jun 26, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/11213 New embedding models like [Qwen3 Embedding](https://qwenlm.github.io/blog/qwen3-embedding/) (e.g. [Qwen3-Embedding-8B](https://huggingface.co/Qwen/Qwen3-Embedding-8B)) has "MRL Support", meaning one can configure the vector dimension for the final embedding. The 8B model, for example, has a default dimension of 4096. This, unfortunately, implies that it cannot be used with pgvector from Postgres, unless Ollama adds a new `dimensions` parameter (like [DeepInfra](https://deepinfra.com/Qwen/Qwen3-Embedding-4B/api?example=http)) for supporting changing the default model dimension. An example usage of this parameter is [here](https://github.com/QwenLM/Qwen3-Embedding/blob/main/examples/qwen3_embedding_transformers.py).
GiteaMirror added the feature request label 2026-05-04 18:07:31 -05:00
GiteaMirror commented 2026-05-04 18:07:33 -05:00
Author
Owner
Copy Link

@drifkin commented on GitHub (Jun 27, 2025):

for MRL models I believe you should be able to manually truncate (at the trained boundaries) and normalize, though that's a bit annoying to do. But a workaround in the meantime

<!-- gh-comment-id:3014645924 --> @drifkin commented on GitHub (Jun 27, 2025): for MRL models I believe you should be able to manually truncate (at the trained boundaries) and normalize, though that's a bit annoying to do. But a workaround in the meantime
GiteaMirror commented 2026-05-04 18:07:33 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 17, 2025):

#12242

<!-- gh-comment-id:3302617029 --> @rick-github commented on GitHub (Sep 17, 2025): #12242
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label feature request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#69445
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?