[GH-ISSUE #4832] llama3:7b cache size set #49566

Closed
opened 2026-04-28 12:16:54 -05:00 by GiteaMirror · 1 comment
Owner

Originally created by @ciscoivan on GitHub (Jun 5, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/4832

1

I installed two NVIDIA RTX 2080 TI graphics cards in an experimental deployment and successfully ran the llama3:7b model. I want to know how to adjust the cache size.

thanks

Originally created by @ciscoivan on GitHub (Jun 5, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/4832 ![1](https://github.com/ollama/ollama/assets/55469637/440bb5f3-d605-4962-a895-3205d7c9d621) I installed two NVIDIA RTX 2080 TI graphics cards in an experimental deployment and successfully ran the llama3:7b model. I want to know how to adjust the cache size. thanks
GiteaMirror added the feature request label 2026-04-28 12:16:54 -05:00
Author
Owner

@jmorganca commented on GitHub (Jun 9, 2024):

Hi @ciscoivan thanks for the issue! What do you mean by cache size? The context window?

If you do mean the context window size, you can do this via the num_ctx option in the API or via /set parameter num_ctx 8192 for example when using ollama run

Hope this helps - if not let me know and I can re-open the issue

<!-- gh-comment-id:2156712354 --> @jmorganca commented on GitHub (Jun 9, 2024): Hi @ciscoivan thanks for the issue! What do you mean by cache size? The context window? If you do mean the context window size, you can do this via the `num_ctx` option in the API or via `/set parameter num_ctx 8192` for example when using `ollama run` Hope this helps - if not let me know and I can re-open the issue
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#49566