github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 08:02:14 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #3159] A way to communicate reasons for low performance to users of CLI & API #27703

New Issue
Closed
opened 2026-04-22 05:14:35 -05:00 by GiteaMirror · 1 comment
GiteaMirror commented 2026-04-22 05:14:35 -05:00
Owner
Copy Link

Originally created by @easp on GitHub (Mar 15, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/3159

Originally assigned to: @bmizerany on GitHub.

People are often concerned about what they perceive to be low performance and/or whether Ollama is making optimal use of their RAM/VRAM/GPU/CPU cores. This comes up frequently in github issues, the main Discord channel and the Discord help channel.

I think a lot of these queries could be avoided if Ollama communicated more information to users in a more concise and obvious way than digging in the log file. Perhaps a message in the CLI, just before displaying the REPL prompt, and a status message field in the API response.

Common conditions seem to be

  • Model + context too large for VRAM
  • GPU detected but not used, using CPU for inference
  • No GPU detected, using CPU for inference

The message should warn that performance will be low and point user towards remedies:

  • Using a model with fewer parameters and/or smaller quantization
  • A URL for troubleshooting issues with GPUs.
Originally created by @easp on GitHub (Mar 15, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/3159 Originally assigned to: @bmizerany on GitHub. People are often concerned about what they perceive to be low performance and/or whether Ollama is making optimal use of their RAM/VRAM/GPU/CPU cores. This comes up frequently in github issues, the main Discord channel and the Discord help channel. I think a lot of these queries could be avoided if Ollama communicated more information to users in a more concise and obvious way than digging in the log file. Perhaps a message in the CLI, just before displaying the REPL prompt, and a status message field in the API response. Common conditions seem to be * Model + context too large for VRAM * GPU detected but not used, using CPU for inference * No GPU detected, using CPU for inference The message should warn that performance will be low and point user towards remedies: * Using a model with fewer parameters and/or smaller quantization * A URL for troubleshooting issues with GPUs.
GiteaMirror added the feature request label 2026-04-22 05:14:36 -05:00
GiteaMirror commented 2026-04-22 05:14:37 -05:00
Author
Owner
Copy Link

@pdevine commented on GitHub (May 18, 2024):

We just added ollama ps in the last version of ollama which will show you what percentage of a model is loaded on the CPU or GPU. For trouble shooting you'll still need to look through the logs, but it should cover the 3 cases you mentioned. I'll go ahead and close out the issue.

<!-- gh-comment-id:2118618471 --> @pdevine commented on GitHub (May 18, 2024): We just added `ollama ps` in the last version of ollama which will show you what percentage of a model is loaded on the CPU or GPU. For trouble shooting you'll still need to look through the logs, but it should cover the 3 cases you mentioned. I'll go ahead and close out the issue.
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 feature request
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#27703
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?