[GH-ISSUE #4303] [BUG]: nomic-embed-text-v1 runs away #2739

Closed
opened 2026-02-22 18:31:01 -05:00 by yindo · 4 comments
Owner

Originally created by @hiiamboris on GitHub (Aug 17, 2025).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/4303

How are you running AnythingLLM?

Docker (remote machine)

What happened?

nomic-embed-text-v1 on some documents eats up all available resources and waits for the end of the world.

Are there known steps to reproduce?

Fresh docker installation on Ubuntu 24 using method 1 from the docs. The server has 4GB of RAM and 6GB of swap.

  1. Set the AI Providers -> Embedder to: AnythingLLM Embedder -> nomic-embed-text-v1
  2. Embed (for example) this 25kB file
  3. After a few minutes it shows this JSON error:
    Image
  4. From now on it just eats up all the rest of the 10GB memory space and deadlocks
Originally created by @hiiamboris on GitHub (Aug 17, 2025). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/4303 ### How are you running AnythingLLM? Docker (remote machine) ### What happened? `nomic-embed-text-v1` on some documents eats up all available resources and waits for the end of the world. ### Are there known steps to reproduce? Fresh docker installation on Ubuntu 24 using [method 1 from the docs](https://github.com/Mintplex-Labs/anything-llm/blob/master/docker/HOW_TO_USE_DOCKER.md). The server has 4GB of RAM and 6GB of swap. 1. Set the AI Providers -> Embedder to: AnythingLLM Embedder -> nomic-embed-text-v1 2. Embed (for example) [this **25kB** file](https://raw.githubusercontent.com/hiiamboris/red/fece179699bd0895b8375affdd0dde11dfbda50a/tests/view-test.red) 3. After a few minutes it shows this JSON error:<br><img width="400" height="1761" alt="Image" src="https://github.com/user-attachments/assets/8426ea73-507f-4356-912d-d170e0917da3" /> 4. From now on it just eats up all the rest of the 10GB memory space and deadlocks
yindo added the possible bug label 2026-02-22 18:31:01 -05:00
yindo closed this issue 2026-02-22 18:31:01 -05:00
Author
Owner

@timothycarambat commented on GitHub (Aug 17, 2025):

nomic-embed-text-v1 on some documents eats up all available resources and waits for the end of the world.

Image

4GB of RAM and 6GB of swap.

Its a larger embedding model and that tradeoff comes with a cost - RAM is likely too low for this to load it into memory and do the chunk sizes provided to it. Provide a machine with more RAM or use the default model - which is smaller and can run on your specs - at the expense of accuracy. Lastly, you can offload embedding to another server or cloud provider if none of the above is possible.

@timothycarambat commented on GitHub (Aug 17, 2025): > nomic-embed-text-v1 on some documents eats up all available resources and waits for the end of the world. <img width="738" height="326" alt="Image" src="https://github.com/user-attachments/assets/48e21592-92af-439d-b7be-8834c5223c6a" /> > 4GB of RAM and 6GB of swap. Its a larger embedding model and that tradeoff comes with a cost - RAM is likely too low for this to load it into memory and do the chunk sizes provided to it. Provide a machine with more RAM or use the default model - which is smaller and can run on your specs - at the expense of accuracy. Lastly, you can offload embedding to another server or cloud provider if none of the above is possible.
Author
Owner

@hiiamboris commented on GitHub (Aug 18, 2025):

Bro. You kidding me? 10GB to split a damn 25kB file? Really? Then why it does work fine for other files, even much bigger ones, and RAM consumption never exceeds 1GB?

@hiiamboris commented on GitHub (Aug 18, 2025): Bro. You kidding me? 10GB to split a damn 25kB file? Really? Then why it does work fine for other files, even much bigger ones, and RAM consumption never exceeds 1GB?
Author
Owner

@hiiamboris commented on GitHub (Aug 18, 2025):

https://huggingface.co/nomic-ai/nomic-embed-text-v1.5/discussions/15 cites up to 524MB RAM requirement.

@hiiamboris commented on GitHub (Aug 18, 2025): https://huggingface.co/nomic-ai/nomic-embed-text-v1.5/discussions/15 cites up to 524MB RAM requirement.
Author
Owner

@hiiamboris commented on GitHub (Aug 20, 2025):

LocalAI handles that file with a v2 (heavier) model within a few seconds. No issues or heavy resource usage.

@hiiamboris commented on GitHub (Aug 20, 2025): LocalAI handles that file with a v2 (heavier) model within a few seconds. No issues or heavy resource usage.
yindo changed title from [BUG]: nomic-embed-text-v1 runs away to [GH-ISSUE #4303] [BUG]: nomic-embed-text-v1 runs away 2026-06-05 14:48:15 -04:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Mintplex-Labs/anything-llm#2739