github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #9616] Performance hit since v0.5.10 #6273

New Issue
Closed
opened 2026-04-12 17:41:52 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-12 17:41:52 -05:00
Owner
Copy Link

Originally created by @webdev23 on GitHub (Mar 9, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/9616

Originally assigned to: @jmorganca on GitHub.

What is the issue?

Clear performance hit, of over 20%, on my machine, between [v0.5.10] and later versions, gradually decreases.

Following multiple benchmarks over multiple versions.

The offloading to CPU is more prominent on later versions.

As a result, on my machine 14B models such as phi4 or deepseek-14B, cannot run at all, with less than 1 token by seconds, leading quickly to CPU over temperature with all cores at 100%, while this behavior never happen on [v0.5.10] .

On [v0.5.10] , on my machine, such 14B could run, at usable speed, 3 or 5 tokens by seconds are doable, over many minutes, not a problem.

Relevant log output

OS

Linux

GPU

Nvidia

CPU

Intel

Ollama version

[v0.5.13]

Originally created by @webdev23 on GitHub (Mar 9, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/9616 Originally assigned to: @jmorganca on GitHub. ### What is the issue? Clear performance hit, of over 20%, on my machine, between [v0.5.10] and later versions, gradually decreases. Following multiple benchmarks over multiple versions. The offloading to CPU is more prominent on later versions. As a result, on my machine 14B models such as phi4 or deepseek-14B, cannot run at all, with less than 1 token by seconds, leading quickly to CPU over temperature with all cores at 100%, while this behavior never happen on [v0.5.10] . On [v0.5.10] , _on my machine_, such 14B could run, at usable speed, 3 or 5 tokens by seconds are doable, over many minutes, not a problem. ### Relevant log output ```shell ``` ### OS Linux ### GPU Nvidia ### CPU Intel ### Ollama version [v0.5.13]
GiteaMirror added the bug label 2026-04-12 17:41:52 -05:00
GiteaMirror commented 2026-04-12 17:41:54 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Mar 9, 2025):

Server logs may aid in debugging.

<!-- gh-comment-id:2709037562 --> @rick-github commented on GitHub (Mar 9, 2025): [Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) may aid in debugging.
GiteaMirror commented 2026-04-12 17:41:54 -05:00
Author
Owner
Copy Link

@jmorganca commented on GitHub (Mar 9, 2025):

@webdev23 sorry about that. As @rick-github mentioned server logs will help us a ton. Ideally for both 0.5.9 and 0.5.13 – thanks so much

<!-- gh-comment-id:2709061633 --> @jmorganca commented on GitHub (Mar 9, 2025): @webdev23 sorry about that. As @rick-github mentioned server logs will help us a ton. Ideally for both 0.5.9 and 0.5.13 – thanks so much
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#6273
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?