github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 00:22:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #467] Running a 70B Model with 16GB RAM: Possible Strategies? #46727

New Issue
Closed
opened 2026-04-27 23:43:57 -05:00 by GiteaMirror · 4 comments
GiteaMirror commented 2026-04-27 23:43:57 -05:00
Owner
Copy Link

Originally created by @OguzcanOzdemir on GitHub (Sep 3, 2023).
Original GitHub issue: https://github.com/ollama/ollama/issues/467

Hello,

I'm currently working with a system that has 16GB of RAM, and I'm interested in running a 70B model for my project. I understand that according to the GitHub repository's documentation, a 70B model typically requires 32GB of RAM.

However, due to my system limitations, I'm looking for guidance on potential strategies or alternative methods to run a 70B model efficiently with 16GB of RAM.

Are there any techniques, optimizations, or workarounds that I can explore to make this possible? I would greatly appreciate any advice or suggestions on how to approach this challenge and still achieve acceptable performance.

Thank you for your assistance and insights.

Originally created by @OguzcanOzdemir on GitHub (Sep 3, 2023). Original GitHub issue: https://github.com/ollama/ollama/issues/467 Hello, I'm currently working with a system that has 16GB of RAM, and I'm interested in running a 70B model for my project. I understand that according to the GitHub repository's documentation, a 70B model typically requires 32GB of RAM. However, due to my system limitations, I'm looking for guidance on potential strategies or alternative methods to run a 70B model efficiently with 16GB of RAM. Are there any techniques, optimizations, or workarounds that I can explore to make this possible? I would greatly appreciate any advice or suggestions on how to approach this challenge and still achieve acceptable performance. Thank you for your assistance and insights.
GiteaMirror added the question label 2026-04-27 23:43:57 -05:00
GiteaMirror commented 2026-04-27 23:44:02 -05:00
Author
Owner
Copy Link

@ddkwork commented on GitHub (Sep 4, 2023):

are you tested 70b？

<!-- gh-comment-id:1705442700 --> @ddkwork commented on GitHub (Sep 4, 2023): are you tested 70b？
GiteaMirror commented 2026-04-27 23:44:09 -05:00
Author
Owner
Copy Link

@OguzcanOzdemir commented on GitHub (Sep 4, 2023):

are you tested 70b？

Yes I did. I got this message "Error: Post "http://127.0.0.1:11434/api/generate": EOF"

<!-- gh-comment-id:1705657340 --> @OguzcanOzdemir commented on GitHub (Sep 4, 2023): > are you tested 70b？ Yes I did. I got this message "Error: Post "http://127.0.0.1:11434/api/generate": EOF"
GiteaMirror commented 2026-04-27 23:44:12 -05:00
Author
Owner
Copy Link

@technovangelist commented on GitHub (Sep 5, 2023):

You got that message because you don't have enough memory. Unfortunately there isn't anything that can be done about that. You really have two strategies available. Use a cloud instance with enough memory, or upgrade to a machine with 64GB of memory. Anything below 64 and you are likely to see that kind of message. Why do you think you need a 70B model? For most cases, smaller models will suffice. Perhaps try some smaller models that will fit on your hardware. I will close this issue, since there is nothing that can be done on our side.

<!-- gh-comment-id:1706908419 --> @technovangelist commented on GitHub (Sep 5, 2023): You got that message because you don't have enough memory. Unfortunately there isn't anything that can be done about that. You really have two strategies available. Use a cloud instance with enough memory, or upgrade to a machine with 64GB of memory. Anything below 64 and you are likely to see that kind of message. Why do you think you need a 70B model? For most cases, smaller models will suffice. Perhaps try some smaller models that will fit on your hardware. I will close this issue, since there is nothing that can be done on our side.
GiteaMirror commented 2026-04-27 23:44:18 -05:00
Author
Owner
Copy Link

@OguzcanOzdemir commented on GitHub (Sep 5, 2023):

You got that message because you don't have enough memory. Unfortunately there isn't anything that can be done about that. You really have two strategies available. Use a cloud instance with enough memory, or upgrade to a machine with 64GB of memory. Anything below 64 and you are likely to see that kind of message. Why do you think you need a 70B model? For most cases, smaller models will suffice. Perhaps try some smaller models that will fit on your hardware. I will close this issue, since there is nothing that can be done on our side.

For some difficult questions, I thought the 70b model would provide more creative answers. Thanks for the answer. I will continue to use the 13b model.

<!-- gh-comment-id:1707218831 --> @OguzcanOzdemir commented on GitHub (Sep 5, 2023): > You got that message because you don't have enough memory. Unfortunately there isn't anything that can be done about that. You really have two strategies available. Use a cloud instance with enough memory, or upgrade to a machine with 64GB of memory. Anything below 64 and you are likely to see that kind of message. Why do you think you need a 70B model? For most cases, smaller models will suffice. Perhaps try some smaller models that will fit on your hardware. I will close this issue, since there is nothing that can be done on our side. For some difficult questions, I thought the 70b model would provide more creative answers. Thanks for the answer. I will continue to use the 13b model.
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 question
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#46727
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?