github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-09 01:59:50 -05:00
Code Issues 14.6k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #3643] how to change the max input token length when I run ‘’ollama run gemma:7b-instruct-v1.1-fp16‘’ #2246

New Issue
Closed
opened 2026-04-12 12:31:14 -05:00 by GiteaMirror · 8 comments
GiteaMirror commented 2026-04-12 12:31:14 -05:00
Owner
Copy Link

Originally created by @dh12306 on GitHub (Apr 15, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/3643

the default input token lens is 2048 ？ how can I change it because the gemma can support more input tokens

Originally created by @dh12306 on GitHub (Apr 15, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/3643 the default input token lens is 2048 ？ how can I change it because the gemma can support more input tokens
GiteaMirror commented 2026-04-12 12:31:15 -05:00
Author
Owner
Copy Link

@dims commented on GitHub (Apr 15, 2024):

tip from @jmorganca here - https://github.com/ollama/ollama/issues/3644#issuecomment-2057646417

<!-- gh-comment-id:2057649187 --> @dims commented on GitHub (Apr 15, 2024): tip from @jmorganca here - https://github.com/ollama/ollama/issues/3644#issuecomment-2057646417
GiteaMirror commented 2026-04-12 12:31:16 -05:00
Author
Owner
Copy Link

@jmorganca commented on GitHub (Apr 17, 2024):

Hi all, to change the max token length you can use /set parameter num_ctx <context size> e.g. 4096, 8192 or more

Hope this helps!

<!-- gh-comment-id:2060137664 --> @jmorganca commented on GitHub (Apr 17, 2024): Hi all, to change the max token length you can use `/set parameter num_ctx <context size>` e.g. `4096`, `8192` or more Hope this helps!
GiteaMirror commented 2026-04-12 12:31:16 -05:00
Author
Owner
Copy Link

@0sengseng0 commented on GitHub (Apr 18, 2024):

Hi all, to change the max token length you can use /set parameter num_ctx <context size> e.g. 4096, 8192 or more大家好，要更改最大令牌长度，您可以使用/set parameter num_ctx <context size>例如4096，8192或更多

Hope this helps! 希望这会有所帮助！

As I increased the token size, it seemed like every time I asked a question, the model would restart before answering。 /set parameter num_ctx 40960

<!-- gh-comment-id:2063563014 --> @0sengseng0 commented on GitHub (Apr 18, 2024): > Hi all, to change the max token length you can use `/set parameter num_ctx <context size>` e.g. `4096`, `8192` or more大家好，要更改最大令牌长度，您可以使用`/set parameter num_ctx <context size>`例如`4096`，`8192`或更多 > > Hope this helps! 希望这会有所帮助！ As I increased the token size, it seemed like every time I asked a question, the model would restart before answering。 /set parameter num_ctx 40960
GiteaMirror commented 2026-04-12 12:31:17 -05:00
Author
Owner
Copy Link

@BoyuanGao commented on GitHub (Aug 1, 2024):

Hi all, to change the max token length you can use /set parameter num_ctx <context size> e.g. 4096, 8192 or more

Hope this helps!

Hi @jmorganca Would you please tell me how to execute this on windows?

<!-- gh-comment-id:2261810477 --> @BoyuanGao commented on GitHub (Aug 1, 2024): > Hi all, to change the max token length you can use `/set parameter num_ctx <context size>` e.g. `4096`, `8192` or more > > Hope this helps! Hi @jmorganca Would you please tell me how to execute this on windows?
GiteaMirror commented 2026-04-12 12:31:17 -05:00
Author
Owner
Copy Link

@vish01 commented on GitHub (Aug 28, 2024):

If you're running this using the Ollama JS package, you can pass it in the options property:

await ollama.generate({
  model: "llama3.1",
  prompt: finalPrompt,
  stream: true,
  options: {
    num_ctx: xx
  },
})
<!-- gh-comment-id:2313875078 --> @vish01 commented on GitHub (Aug 28, 2024): If you're running this using the Ollama JS package, you can pass it in the `options` property: ``` await ollama.generate({ model: "llama3.1", prompt: finalPrompt, stream: true, options: { num_ctx: xx }, }) ```
GiteaMirror commented 2026-04-12 12:31:17 -05:00
Author
Owner
Copy Link

@vitustockholm commented on GitHub (Aug 30, 2024):

Hi all, to change the max token length you can use /set parameter num_ctx <context size> e.g. 4096, 8192 or more
Hope this helps!

Hi @jmorganca Would you please tell me how to execute this on windows?

use webui to change this parameter

<!-- gh-comment-id:2322122472 --> @vitustockholm commented on GitHub (Aug 30, 2024): > > Hi all, to change the max token length you can use `/set parameter num_ctx <context size>` e.g. `4096`, `8192` or more > > Hope this helps! > > Hi @jmorganca Would you please tell me how to execute this on windows? use webui to change this parameter
GiteaMirror commented 2026-04-12 12:31:17 -05:00
Author
Owner
Copy Link

@Arslan-Mehmood1 commented on GitHub (Dec 2, 2024):

as I faced this issue while using ollama with langchain, I checked the api reference for loading model and found then I can set parameter 'num_ctx' for context window length.

https://python.langchain.com/api_reference/ollama/llms/langchain_ollama.llms.OllamaLLM.html#langchain_ollama.llms.OllamaLLM.num_ctx

<!-- gh-comment-id:2511065858 --> @Arslan-Mehmood1 commented on GitHub (Dec 2, 2024): as I faced this issue while using ollama with langchain, I checked the api reference for loading model and found then I can set parameter 'num_ctx' for context window length. https://python.langchain.com/api_reference/ollama/llms/langchain_ollama.llms.OllamaLLM.html#langchain_ollama.llms.OllamaLLM.num_ctx
GiteaMirror commented 2026-04-12 12:31:18 -05:00
Author
Owner
Copy Link

@PasaOpasen commented on GitHub (Nov 12, 2025):

is there a way to globally limit this value from environment ?

<!-- gh-comment-id:3520924289 --> @PasaOpasen commented on GitHub (Nov 12, 2025): is there a way to globally limit this value from environment ?
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
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#2246
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?