[GH-ISSUE #1632] Only utilizing one thread - Unraid #26670

Closed
opened 2026-04-22 03:05:32 -05:00 by GiteaMirror · 1 comment
Owner

Originally created by @evanrodgers on GitHub (Dec 20, 2023).
Original GitHub issue: https://github.com/ollama/ollama/issues/1632

Hi all,

Right now I am using Ollama 0.1.17 in an Unraid environment. The app works great, but it's only utilizing one thread. I did some light troubleshooting by adding the "num_thread" parameter to the modelfile as shown below, but it still only utilizes one thread. I've checked this by looking at the dashboard in Unraid and HTOP.

Screenshot 2023-12-20 095344

Modelfile:

# Modelfile generated by "ollama show"
# To build a new Modelfile based on this one, replace the FROM line with:
# FROM solar:10.7b-instruct-v1-q4_K_S

FROM /root/.ollama/models/blobs/sha256:034d7d28eda7c13de702e09978ffbc85d8fcc0ac173e5ecb4d0c626507fb25b6
TEMPLATE """### System:
{{ .System }}

### User:
{{ .Prompt }}

### Assistant:
"""
PARAMETER num_ctx 4096
PARAMETER num_thread 6
PARAMETER stop "</s>"
PARAMETER stop "### System:"
PARAMETER stop "### User:"
PARAMETER stop "### Assistant:"

Thanks everyone for all your great work!

EDIT: I was able to fix this, but to be honest I'm unclear what the solution was. I rebooted (always good), disabled SMT, disabled CSM, enabled 4G Decoding, enabled Resizable Bar. Sorry for the trouble!

Originally created by @evanrodgers on GitHub (Dec 20, 2023). Original GitHub issue: https://github.com/ollama/ollama/issues/1632 Hi all, Right now I am using Ollama 0.1.17 in an Unraid environment. The app works great, but it's only utilizing one thread. I did some light troubleshooting by adding the "num_thread" parameter to the modelfile as shown below, but it still only utilizes one thread. I've checked this by looking at the dashboard in Unraid and HTOP. ![Screenshot 2023-12-20 095344](https://github.com/jmorganca/ollama/assets/36175609/4b2451c4-26d5-4f52-b633-95cc0248fed8) Modelfile: ``` # Modelfile generated by "ollama show" # To build a new Modelfile based on this one, replace the FROM line with: # FROM solar:10.7b-instruct-v1-q4_K_S FROM /root/.ollama/models/blobs/sha256:034d7d28eda7c13de702e09978ffbc85d8fcc0ac173e5ecb4d0c626507fb25b6 TEMPLATE """### System: {{ .System }} ### User: {{ .Prompt }} ### Assistant: """ PARAMETER num_ctx 4096 PARAMETER num_thread 6 PARAMETER stop "</s>" PARAMETER stop "### System:" PARAMETER stop "### User:" PARAMETER stop "### Assistant:" ``` Thanks everyone for all your great work! **EDIT: I was able to fix this**, but to be honest I'm unclear what the solution was. I rebooted (always good), disabled SMT, disabled CSM, enabled 4G Decoding, enabled Resizable Bar. Sorry for the trouble!
Author
Owner

@mchiang0610 commented on GitHub (Mar 11, 2024):

@evanrodgers thank you for sharing. Sorry about that. This should be fixed in subsequent versions. Please let us know if a regression ever happens. Again, sorry about this.

<!-- gh-comment-id:1989060899 --> @mchiang0610 commented on GitHub (Mar 11, 2024): @evanrodgers thank you for sharing. Sorry about that. This should be fixed in subsequent versions. Please let us know if a regression ever happens. Again, sorry about this.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#26670