github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-08 00:51:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[PR #3092] [CLOSED] Limit num_predict to num_ctx #16335

New Issue
Closed
opened 2026-04-16 05:25:29 -05:00 by GiteaMirror · 0 comments
GiteaMirror commented 2026-04-16 05:25:29 -05:00
Owner
Copy Link

📋 Pull Request Information

Original PR: https://github.com/ollama/ollama/pull/3092
Author: @jmorganca
Created: 3/13/2024
Status: ❌ Closed

Base: main ← Head: jmorganca/limit

📝 Commits (1)

  • ca7c3f7 limit num_predict to num_ctx

📊 Changes

1 file changed (+13 additions, -0 deletions)

View changed files

📝 llm/dyn_ext_server.go (+13 -0)

📄 Description

This limits the number of tokens generated to the context window size, allowing a maximum of two "context shifts" should the context window limit be passed. It will also provide a maximum token limit to stop any "runaway" prompts that occur from smaller models that continue to generate indefinitely (e.g. in JSON mode)

Ideally we have no context shifts: we only generate the number of tokens left in the context window after the prompt, but this felt like a simple change that would ease our way into this since chat prompts can get quite large and we'd have to change our prompt trimming strategy to leave enough space for longer responses.

🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.

## 📋 Pull Request Information **Original PR:** https://github.com/ollama/ollama/pull/3092 **Author:** [@jmorganca](https://github.com/jmorganca) **Created:** 3/13/2024 **Status:** ❌ Closed **Base:** `main` ← **Head:** `jmorganca/limit` --- ### 📝 Commits (1) - [`ca7c3f7`](https://github.com/ollama/ollama/commit/ca7c3f7e0f9b14ea435aedef1c448a37a44f423c) limit `num_predict` to `num_ctx` ### 📊 Changes **1 file changed** (+13 additions, -0 deletions) <details> <summary>View changed files</summary> 📝 `llm/dyn_ext_server.go` (+13 -0) </details> ### 📄 Description This limits the number of tokens generated to the context window size, allowing a maximum of two "context shifts" should the context window limit be passed. It will also provide a maximum token limit to stop any "runaway" prompts that occur from smaller models that continue to generate indefinitely (e.g. in JSON mode) Ideally we have no context shifts: we only generate the number of tokens left in the context window after the prompt, but this felt like a simple change that would ease our way into this since chat prompts can get quite large and we'd have to change our prompt trimming strategy to leave enough space for longer responses. --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>
GiteaMirror added the pull-request label 2026-04-16 05:25:29 -05:00
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 pull-request
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#16335
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?