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 #5680] Extremely slow on Mac M1 chip #29301

New Issue
Closed
opened 2026-04-22 08:03:22 -05:00 by GiteaMirror · 7 comments
GiteaMirror commented 2026-04-22 08:03:22 -05:00
Owner
Copy Link

Originally created by @lulunac27a on GitHub (Jul 13, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/5680

Originally assigned to: @jmorganca on GitHub.

What is the issue?

I tried chatting using Llama from Meta AI, when the answer is generating, my computer is so slow and sometimes freezes (like my mouse not moving when I move the trackpad). It takes few minutes to completely generate an answer from a question. I use Apple M1 chip with 8GB of RAM memory.

OS

macOS

GPU

Apple

CPU

Apple

Ollama version

0.2.2

Originally created by @lulunac27a on GitHub (Jul 13, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/5680 Originally assigned to: @jmorganca on GitHub. ### What is the issue? I tried chatting using Llama from Meta AI, when the answer is generating, my computer is so slow and sometimes freezes (like my mouse not moving when I move the trackpad). It takes few minutes to completely generate an answer from a question. I use Apple M1 chip with 8GB of RAM memory. ### OS macOS ### GPU Apple ### CPU Apple ### Ollama version 0.2.2
GiteaMirror added the bug label 2026-04-22 08:03:22 -05:00
GiteaMirror commented 2026-04-22 08:03:26 -05:00
Author
Owner
Copy Link

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

What was the size of the model? Most models are multi-gigabytes in size, if you tried to run one of these with 8G of RAM then I imagine that it will cause a lot of swap activity. Try running a smaller model (eg, https://ollama.com/library/qwen2:0.5b) and see if performance improves.

<!-- gh-comment-id:2227099513 --> @rick-github commented on GitHub (Jul 13, 2024): What was the size of the model? Most models are multi-gigabytes in size, if you tried to run one of these with 8G of RAM then I imagine that it will cause a lot of swap activity. Try running a smaller model (eg, https://ollama.com/library/qwen2:0.5b) and see if performance improves.
GiteaMirror commented 2026-04-22 08:03:27 -05:00
Author
Owner
Copy Link

@lulunac27a commented on GitHub (Jul 14, 2024):

llama3 with 8B size

<!-- gh-comment-id:2227157868 --> @lulunac27a commented on GitHub (Jul 14, 2024): llama3 with 8B size
GiteaMirror commented 2026-04-22 08:03:28 -05:00
Author
Owner
Copy Link

@sudochia commented on GitHub (Jul 14, 2024):

Try https://ollama.com/library/tinyllama and see how it goes?

<!-- gh-comment-id:2227208562 --> @sudochia commented on GitHub (Jul 14, 2024): Try https://ollama.com/library/tinyllama and see how it goes?
GiteaMirror commented 2026-04-22 08:03:29 -05:00
Author
Owner
Copy Link

@igorschlum commented on GitHub (Jul 15, 2024):

@lulunac27a you will need a 16GB mac, but to learn Tinyllama is good. You can also use llama3:8b-instruct-q2_K and restart your Mac before and use few apps to let max amount of memory to Ollama and the model.

<!-- gh-comment-id:2227661021 --> @igorschlum commented on GitHub (Jul 15, 2024): @lulunac27a you will need a 16GB mac, but to learn Tinyllama is good. You can also use llama3:8b-instruct-q2_K and restart your Mac before and use few apps to let max amount of memory to Ollama and the model.
GiteaMirror commented 2026-04-22 08:03:30 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Jul 23, 2024):

As others have pointed out, llama3 is going to be large for your system if you have anything else using memory.

> ollama ps
NAME         	ID          	SIZE  	PROCESSOR	UNTIL
llama3:latest	a6990ed6be41	5.5 GB	100% GPU 	4 minutes from now

On an 8G Mac, only ~5.7GB is available for VRAM usage, not leaving much buffer, so while it does work, it is pushing the limits of your system.

<!-- gh-comment-id:2246293807 --> @dhiltgen commented on GitHub (Jul 23, 2024): As others have pointed out, llama3 is going to be large for your system if you have anything else using memory. ``` > ollama ps NAME ID SIZE PROCESSOR UNTIL llama3:latest a6990ed6be41 5.5 GB 100% GPU 4 minutes from now ``` On an 8G Mac, only ~5.7GB is available for VRAM usage, not leaving much buffer, so while it does work, it is pushing the limits of your system.
GiteaMirror commented 2026-04-22 08:03:30 -05:00
Author
Owner
Copy Link

@piotrszczesniak commented on GitHub (Sep 25, 2024):

Try https://ollama.com/library/tinyllama and see how it goes?

@sudochia - this model is way quicker to respond, thanks!

<!-- gh-comment-id:2375199305 --> @piotrszczesniak commented on GitHub (Sep 25, 2024): > Try https://ollama.com/library/tinyllama and see how it goes? @sudochia - this model is way quicker to respond, thanks!
GiteaMirror commented 2026-04-22 08:03:32 -05:00
Author
Owner
Copy Link

@igorschlum commented on GitHub (Sep 26, 2024):

Hi @piotrszczesniak, you can try pulling llama3.2 which is a smaller LLM and may respond better than tinyllama.

<!-- gh-comment-id:2377017426 --> @igorschlum commented on GitHub (Sep 26, 2024): Hi @piotrszczesniak, you can try pulling [llama3.2](https://ollama.com/library/llama3.2) which is a smaller LLM and may respond better than tinyllama.
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#29301
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?