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 #8357] Faiss document dependency error #67415

New Issue
Closed
opened 2026-05-04 10:17:09 -05:00 by GiteaMirror · 6 comments
GiteaMirror commented 2026-05-04 10:17:09 -05:00
Owner
Copy Link

Originally created by @suckseed5 on GitHub (Jan 9, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/8357

What is the issue?

Hello, we have some questions about similar document matching and look forward to your reply.

image
Ollama Embeddings Model: quentinz/bge-large-zh-v1.5:latest
Vector Stores: faiss
Index: Some knowledge base about legal documents
Ask Question: What's the weather like today?

Why is the question completely unrelated to the vector library, but it still matches the content and has a high score?

OS

No response

GPU

No response

CPU

No response

Ollama version

No response

Originally created by @suckseed5 on GitHub (Jan 9, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/8357 ### What is the issue? Hello, we have some questions about similar document matching and look forward to your reply. ![image](https://github.com/user-attachments/assets/529de6eb-bc87-4ac6-942d-49f75feb1744) Ollama Embeddings Model: quentinz/bge-large-zh-v1.5:latest Vector Stores: faiss Index: Some knowledge base about legal documents Ask Question: What's the weather like today? Why is the question completely unrelated to the vector library, but it still matches the content and has a high score? ### OS _No response_ ### GPU _No response_ ### CPU _No response_ ### Ollama version _No response_
GiteaMirror added the question label 2026-05-04 10:17:09 -05:00
GiteaMirror commented 2026-05-04 10:17:20 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Jan 9, 2025):

Seems like a question for whoever wrote the scoring function.

<!-- gh-comment-id:2579258231 --> @rick-github commented on GitHub (Jan 9, 2025): Seems like a question for whoever wrote the scoring function.
GiteaMirror commented 2026-05-04 10:17:25 -05:00
Author
Owner
Copy Link

@suckseed5 commented on GitHub (Jan 9, 2025):

Seems like a question for whoever wrote the scoring function.

Yes, this will directly lead to matching the wrong knowledge base information.

<!-- gh-comment-id:2579285132 --> @suckseed5 commented on GitHub (Jan 9, 2025): > Seems like a question for whoever wrote the scoring function. Yes, this will directly lead to matching the wrong knowledge base information.
GiteaMirror commented 2026-05-04 10:17:27 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Jan 9, 2025):

All ollama does is provide the embedding, questions about the scoring function need to be asked in the appropriate forum. Perhaps if you provide some more information (client, framework, backend, etc), somebody could provide a pointer.

<!-- gh-comment-id:2579302352 --> @rick-github commented on GitHub (Jan 9, 2025): All ollama does is provide the embedding, questions about the scoring function need to be asked in the appropriate forum. Perhaps if you provide some more information (client, framework, backend, etc), somebody could provide a pointer.
GiteaMirror commented 2026-05-04 10:17:32 -05:00
Author
Owner
Copy Link

@suckseed5 commented on GitHub (Jan 9, 2025):

Just replacing the Ollama Embeddings component with HuggingFace Embeddings, it worked. Is Ollama Embeddings inaccurate?
But are there any offline embedding tools?

<!-- gh-comment-id:2579361696 --> @suckseed5 commented on GitHub (Jan 9, 2025): Just replacing the Ollama Embeddings component with HuggingFace Embeddings, it worked. Is Ollama Embeddings inaccurate? But are there any offline embedding tools?
GiteaMirror commented 2026-05-04 10:17:35 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Jan 9, 2025):

Have you tried different embedding models? Perhaps quentinz/bge-large-zh-v1.5 outputs embeddings that are too similar. Note that quentinz/bge-large-zh-v1.5 has a context length of 512 and your chunk size is 1000, so if your input is token-rich, you will be losing semantic information from the embeddings. There are many offline embedding tools, see for example fastembed.

<!-- gh-comment-id:2579445530 --> @rick-github commented on GitHub (Jan 9, 2025): Have you tried different [embedding](https://ollama.com/search?c=embedding) models? Perhaps quentinz/bge-large-zh-v1.5 outputs embeddings that are too similar. Note that quentinz/bge-large-zh-v1.5 has a context length of 512 and your chunk size is 1000, so if your input is token-rich, you will be losing semantic information from the embeddings. There are many offline embedding tools, see for example [fastembed](https://github.com/qdrant/fastembed).
GiteaMirror commented 2026-05-04 10:17:37 -05:00
Author
Owner
Copy Link

@pdevine commented on GitHub (Jan 11, 2025):

I'm going to go ahead and close the issue. It looks like from the model page there is a GH repo. Best to check with the authors there.

<!-- gh-comment-id:2584936284 --> @pdevine commented on GitHub (Jan 11, 2025): I'm going to go ahead and close the issue. It looks like from the [model page](https://github.com/FlagOpen/FlagEmbedding) there is a GH repo. Best to check with the authors there.
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 question
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#67415
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?