github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #651] Ability to configure embeddings dimension size #62327

New Issue
Closed
opened 2026-05-03 08:14:17 -05:00 by GiteaMirror · 9 comments
GiteaMirror commented 2026-05-03 08:14:17 -05:00
Owner
Copy Link

Originally created by @jamesbraza on GitHub (Sep 29, 2023).
Original GitHub issue: https://github.com/ollama/ollama/issues/651

Moving from OllamaEmbeddings(model="llama2:13b") to OllamaEmbeddings(model="llama2:7b"), I am now getting a shape mismatch in my embeddings:

ValueError: shapes (4096,) and (5120,) not aligned: 4096 (dim 0) != 5120 (dim 0)

So the 7b embeddings is slightly smaller (4096) than 13b embeddings (5120). Is there an argument or parameter I can use to control the embedding size?

I would like to artificially switch 7b to use 5120, so I can not rebuild my vector store.

Originally created by @jamesbraza on GitHub (Sep 29, 2023). Original GitHub issue: https://github.com/ollama/ollama/issues/651 Moving from `OllamaEmbeddings(model="llama2:13b")` to `OllamaEmbeddings(model="llama2:7b")`, I am now getting a shape mismatch in my embeddings: ```none ValueError: shapes (4096,) and (5120,) not aligned: 4096 (dim 0) != 5120 (dim 0) ``` So the `7b` embeddings is slightly smaller (4096) than `13b` embeddings (5120). Is there an argument or parameter I can use to control the embedding size? I would like to artificially switch `7b` to use 5120, so I can not rebuild my vector store.
GiteaMirror added the embeddingsfeature request labels 2026-05-03 08:14:17 -05:00
GiteaMirror commented 2026-05-03 08:14:38 -05:00
Author
Owner
Copy Link

@kittycattoys commented on GitHub (Feb 22, 2024):

This functionality from Nomic is nice. I am using TinyLlama / Llama2 for embeddings but would like the option to change dimensionality like Nomic has: from nomic import embed response = embed.text(
texts=[prompt],
model='nomic-embed-text-v1.5',
task_type='search_document',
dimensionality=768
) Is this enhancement possible?

<!-- gh-comment-id:1960501880 --> @kittycattoys commented on GitHub (Feb 22, 2024): This functionality from Nomic is nice. I am using TinyLlama / Llama2 for embeddings but would like the option to change dimensionality like Nomic has: from nomic import embed response = embed.text( texts=[prompt], model='nomic-embed-text-v1.5', task_type='search_document', dimensionality=768 ) Is this enhancement possible?
GiteaMirror commented 2026-05-03 08:14:40 -05:00
Author
Owner
Copy Link

@bilunsun commented on GitHub (Mar 6, 2024):

I second this.

This functionality from Nomic is nice. I am using TinyLlama / Llama2 for embeddings but would like the option to change dimensionality like Nomic has: from nomic import embed response = embed.text( texts=[prompt], model='nomic-embed-text-v1.5', task_type='search_document', dimensionality=768 ) Is this enhancement possible?

In their latest v1.5 HuggingFace page, the user should be able to configure both the task_type and dimensionality. However, I have not been able to set these arguments with Ollama.

<!-- gh-comment-id:1981286906 --> @bilunsun commented on GitHub (Mar 6, 2024): I second this. > This functionality from Nomic is nice. I am using TinyLlama / Llama2 for embeddings but would like the option to change dimensionality like Nomic has: from nomic import embed response = embed.text( texts=[prompt], model='nomic-embed-text-v1.5', task_type='search_document', dimensionality=768 ) Is this enhancement possible? In their latest v1.5 [HuggingFace page](https://huggingface.co/nomic-ai/nomic-embed-text-v1.5), the user should be able to configure both the `task_type` and `dimensionality`. However, I have not been able to set these arguments with Ollama.
GiteaMirror commented 2026-05-03 08:14:41 -05:00
Author
Owner
Copy Link

@kenneth-liao commented on GitHub (Jul 25, 2024):

Agree, this type of functionality is critical for optimizing the embedding process and database size when storing embeddings.

<!-- gh-comment-id:2250975996 --> @kenneth-liao commented on GitHub (Jul 25, 2024): Agree, this type of functionality is critical for optimizing the embedding process and database size when storing embeddings.
GiteaMirror commented 2026-05-03 08:14:42 -05:00
Author
Owner
Copy Link

@raffaeler commented on GitHub (Nov 19, 2024):

FYI if the model has Matryoshka capabilities, cutting the vector to the desired size is straightforward. You just keep the most significant portion:

  1. take the [0:length] part
  2. normalize the result vector
<!-- gh-comment-id:2486161758 --> @raffaeler commented on GitHub (Nov 19, 2024): FYI if the model has Matryoshka capabilities, cutting the vector to the desired size is straightforward. You just keep the most significant portion: 1. take the [0:length] part 2. normalize the result vector
GiteaMirror commented 2026-05-03 08:14:44 -05:00
Author
Owner
Copy Link

@saadshaikh3 commented on GitHub (Jan 17, 2025):

Any update on this feature ?

<!-- gh-comment-id:2598560856 --> @saadshaikh3 commented on GitHub (Jan 17, 2025): Any update on this feature ?
GiteaMirror commented 2026-05-03 08:14:48 -05:00
Author
Owner
Copy Link

@noahfraiture commented on GitHub (Mar 8, 2025):

Still nothing ?

<!-- gh-comment-id:2708354011 --> @noahfraiture commented on GitHub (Mar 8, 2025): Still nothing ?
GiteaMirror commented 2026-05-03 08:14:48 -05:00
Author
Owner
Copy Link

@PhilippeCalvez commented on GitHub (Mar 17, 2025):

On my side the problem was due to a rate limited litellm proxy that I'm using. I've installed the model nomic-embed-text on a local ollama server. It works but I had to specify the embedding dimension.

# Define the Crew crew = Crew( agents=agents, tasks=tasks, process=Process.sequential, verbose=True, memory=True, embedder={ "provider": "ollama", "config": { "model": 'nomic-embed-text', "api_base": "http://localhost:11434/api/embed" } }, embedding_dimension=768 )

<!-- gh-comment-id:2729449070 --> @PhilippeCalvez commented on GitHub (Mar 17, 2025): On my side the problem was due to a rate limited litellm proxy that I'm using. I've installed the model nomic-embed-text on a local ollama server. It works but I had to specify the embedding dimension. `# Define the Crew crew = Crew( agents=agents, tasks=tasks, process=Process.sequential, verbose=True, memory=True, embedder={ "provider": "ollama", "config": { "model": 'nomic-embed-text', "api_base": "http://localhost:11434/api/embed" } }, embedding_dimension=768 ) `
GiteaMirror commented 2026-05-03 08:14:49 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 16, 2025):

https://github.com/ollama/ollama/pull/12242

<!-- gh-comment-id:3298009014 --> @rick-github commented on GitHub (Sep 16, 2025): https://github.com/ollama/ollama/pull/12242
GiteaMirror commented 2026-05-03 08:14:50 -05:00
Author
Owner
Copy Link

@jamesbraza commented on GitHub (Sep 16, 2025):

@rick-github feel free to close this issue if you think that PR resolves it. I made this request two years ago, I don't really have the code to test this any more

<!-- gh-comment-id:3299672921 --> @jamesbraza commented on GitHub (Sep 16, 2025): @rick-github feel free to close this issue if you think that PR resolves it. I made this request two years ago, I don't really have the code to test this any more
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 embeddings 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#62327
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?