Mintplex-Labs/anything-llm
1
0
Fork 0
mirror of https://github.com/Mintplex-Labs/anything-llm.git synced 2026-07-19 14:13:35 -04:00
Code Issues 328 Packages Projects Releases 10 Wiki Activity

[GH-ISSUE #917] Query mode not working #550

New Issue
Closed
opened 2026-02-22 18:20:07 -05:00 by yindo · 1 comment
yindo commented 2026-02-22 18:20:07 -05:00
Owner
Copy Link
Copy Source

Originally created by @ghost on GitHub (Mar 17, 2024).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/917

Using anything llm desktop and LM studio on M2 Mac. Able to create embeddings and vectors but in query mode get a message saying no information in workspace to answer the question. Any help is appreciated.

Originally created by @ghost on GitHub (Mar 17, 2024). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/917 Using anything llm desktop and LM studio on M2 Mac. Able to create embeddings and vectors but in query mode get a message saying no information in workspace to answer the question. Any help is appreciated.
yindo closed this issue 2026-02-22 18:20:07 -05:00
yindo commented 2026-02-22 18:20:08 -05:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (Mar 18, 2024):

This is expected when you send a query that has no text chunk results that have a match over the workspace threshold. Embedding and vector-database semantic searching is not infallible. What is happening here is that the vector db is not finding and context to work with because the prompt, once embedded, is too "far" from other vectors - resulting in a low match.

We give you the controls to change this for any reason. Open the workspace's settings and change the "Threshold similarity" to "No Restriction", and even increase the "Context snippets" from 4-6.

What this does is apply no score-filtering on results from the vector db. It is totally use-case and document dependent on if you need to do this, which is why we expose the controls for these settings.

If you still have poor results, you can even "pin" a document in a workspace, which will do full-text injection in the prompt.

@timothycarambat commented on GitHub (Mar 18, 2024): This is expected when you send a query that has no text chunk results that have a match over the workspace threshold. Embedding and vector-database semantic searching is not infallible. What is happening here is that the vector db is not finding and context to work with because the prompt, once embedded, is too "far" from other vectors - resulting in a low match. We give you the controls to change this for any reason. Open the workspace's settings and change the "Threshold similarity" to "No Restriction", and even increase the "Context snippets" from 4-6. What this does is apply no score-filtering on results from the vector db. It is totally use-case and document dependent on if you need to do this, which is why we expose the controls for these settings. If you still have poor results, you can even "pin" a document in a workspace, which will do full-text injection in the prompt.
yindo changed title from Query mode not working to [GH-ISSUE #917] Query mode not working 2026-06-05 14:35:56 -04:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
Desktop
Docker
Integration Request
Integration Request
OS: Linux
OS: Mobile
OS: Windows
UI/UX
blocked
bug
bug
core-team-only
documentation
duplicate
embed-widget
enhancement
feature request
github_actions
good first issue
investigating
needs info / can't replicate
possible bug
pull-request
Mirrored from GitHub Pull Request
 question
stage: specifications
wontfix
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Mintplex-Labs/anything-llm#550
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?