github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 08:30:05 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #6792] The system parameter OLLAMA_NUM_PALLEL is invalid for embeding model #66319

New Issue
Closed
opened 2026-05-04 02:30:58 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-05-04 02:30:58 -05:00
Owner
Copy Link

Originally created by @black-fox-user on GitHub (Sep 13, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/6792

What is the issue?

I have set the system parameters, but when loading the embedding model, only one is still in effect. I copied this model, and surprisingly, their model IDs are the same. After importing the model, the model ID changed, but the same model was still used in the end。
image

OS

Windows

GPU

Nvidia

CPU

Intel

Ollama version

0.3.10

Originally created by @black-fox-user on GitHub (Sep 13, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/6792 ### What is the issue? I have set the system parameters, but when loading the embedding model, only one is still in effect. I copied this model, and surprisingly, their model IDs are the same. After importing the model, the model ID changed, but the same model was still used in the end。 ![image](https://github.com/user-attachments/assets/e331defe-5548-479b-a5de-32a3d8d1de4d) ### OS Windows ### GPU Nvidia ### CPU Intel ### Ollama version 0.3.10
GiteaMirror added the bug label 2026-05-04 02:30:58 -05:00
GiteaMirror commented 2026-05-04 02:31:12 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 13, 2024):

This is due to a change that was introduced to fix memory corruption in embedding models.

<!-- gh-comment-id:2348870551 --> @rick-github commented on GitHub (Sep 13, 2024): This is due to a [change](https://github.com/ollama/ollama/pull/6467) that was introduced to fix memory corruption in embedding models.
GiteaMirror commented 2026-05-04 02:31:20 -05:00
Author
Owner
Copy Link

@black-fox-user commented on GitHub (Sep 14, 2024):

This is due to a change that was introduced to fix memory corruption in embedding models.
Perhaps my description is not comprehensive. I originally intended to set up the vector model to support concurrency, but only one concurrent number is effective. Then I remembered the OLLAMA_MAX_LOADEDMODELS parameter and copied a model, but with the same ID, the same model was still loaded. I tried importing another model again, although the IDs are different, their digest parameter points to the same model. The OLLAMA_NUMPARALLEL and OLLAMA_MAX_LOADEDMODELS parameters are valid for other models, such as qwen2:1.5b.
1726219316562

<!-- gh-comment-id:2350765281 --> @black-fox-user commented on GitHub (Sep 14, 2024): > This is due to a [change](https://github.com/ollama/ollama/pull/6467) that was introduced to fix memory corruption in embedding models. Perhaps my description is not comprehensive. I originally intended to set up the vector model to support concurrency, but only one concurrent number is effective. Then I remembered the OLLAMA_MAX_LOADEDMODELS parameter and copied a model, but with the same ID, the same model was still loaded. I tried importing another model again, although the IDs are different, their digest parameter points to the same model. The OLLAMA_NUMPARALLEL and OLLAMA_MAX_LOADEDMODELS parameters are valid for other models, such as qwen2:1.5b. ![1726219316562](https://github.com/user-attachments/assets/028cda9a-36ec-4187-99cb-8c6e3075fdb2)
GiteaMirror commented 2026-05-04 02:31:25 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 14, 2024):

ollama doesn't support loading the same model more than once. A model is identified by the sha256 of the GGUF file, so creating a new model using the same GGUF file is just like giving the same model a new name, you still can't load it more than once. If you want to run the same model more than once, you need to run multiple ollama servers.

<!-- gh-comment-id:2350771796 --> @rick-github commented on GitHub (Sep 14, 2024): ollama doesn't support loading the same model more than once. A model is identified by the sha256 of the GGUF file, so creating a new model using the same GGUF file is just like giving the same model a new name, you still can't load it more than once. If you want to run the same model more than once, you need to run multiple ollama servers.
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 bug
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#66319
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?