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

[GH-ISSUE #8732] RFE: Please do not remove the generate context parameter #5665

New Issue
Closed
opened 2026-04-12 16:57:43 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-04-12 16:57:43 -05:00
Owner
Copy Link

Originally created by @asterbini on GitHub (Jan 31, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/8732

I am using the context returned from the generate call to continue the chat without re-interpreting the earlier prompts.
Why this parameter has been deprecated?
Please keep it

Originally created by @asterbini on GitHub (Jan 31, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/8732 I am using the context returned from the generate call to continue the chat without re-interpreting the earlier prompts. Why this parameter has been deprecated? Please keep it
GiteaMirror added the feature request label 2026-04-12 16:57:43 -05:00
GiteaMirror commented 2026-04-12 16:57:45 -05:00
Author
Owner
Copy Link

@meltyli commented on GitHub (Feb 27, 2025):

Is there a substitute option in perhaps the chat endpoint? It is currently more efficient than adding the earlier prompt as part of a following prompt.

<!-- gh-comment-id:2686913777 --> @meltyli commented on GitHub (Feb 27, 2025): Is there a substitute option in perhaps the chat endpoint? It is currently more efficient than adding the earlier prompt as part of a following prompt.
GiteaMirror commented 2026-04-12 16:57:45 -05:00
Author
Owner
Copy Link

@asterbini commented on GitHub (Feb 27, 2025):

As I understand, OpenAI caches the earlier chats so that a list of earlier prompt allows them to internally retrieve the context and continue the chat.
The context parameter allows us to implement a simple, stateless "continue chatting" multi prompt software

<!-- gh-comment-id:2687150418 --> @asterbini commented on GitHub (Feb 27, 2025): As I understand, OpenAI caches the earlier chats so that a list of earlier prompt allows them to internally retrieve the context and continue the chat. The context parameter allows us to implement a simple, stateless "continue chatting" multi prompt software
GiteaMirror commented 2026-04-12 16:57:46 -05:00
Author
Owner
Copy Link

@WizardMiner commented on GitHub (May 5, 2025):

Please do not remove the context parameter.

This is necessary for resuming from prior points in the conversation. For instance, if we want to try a prompt 3 different ways. Or if we want to back up several prompt/response iterations and go different directions. Context seeds the response with where we are now. Without it, we can't roll back the conversation. With context, we can store and reuse this point of view for unlimited kinds of interactions provided the underlying LLM hasn't changed.

Consider if we're in a dam water monitoring scenario, or patient medical health or fraud analysis. We load up all the safeguards into a context, then have the LLM evaluate the prompt within current status. We don't want to have to reload the rules each time or manage context degradation over time. With the context array parameter, we can hit it at that moment of understanding as many times as we want.

Please do not remove the context parameter or give us an alternative to manage our own. The LLMs don't care how we call them. They don't need to have an exclusive on-going chat memory. Using API with the same context array over and over is very useful. Without this feature, energy utilization increases and quality of response decreases. Often LLM come to different conclusions with the same input so that it sounds conversational. This is a huge problem trying to get back to the same state of understanding. Context parameter is critical. Please keep it.

<!-- gh-comment-id:2851953964 --> @WizardMiner commented on GitHub (May 5, 2025): Please do not remove the context parameter. This is necessary for resuming from prior points in the conversation. For instance, if we want to try a prompt 3 different ways. Or if we want to back up several prompt/response iterations and go different directions. Context seeds the response with where we are now. Without it, we can't roll back the conversation. With context, we can store and reuse this point of view for unlimited kinds of interactions provided the underlying LLM hasn't changed. Consider if we're in a dam water monitoring scenario, or patient medical health or fraud analysis. We load up all the safeguards into a context, then have the LLM evaluate the prompt within current status. We don't want to have to reload the rules each time or manage context degradation over time. With the context array parameter, we can hit it at that moment of understanding as many times as we want. Please do not remove the context parameter or give us an alternative to manage our own. The LLMs don't care how we call them. They don't need to have an **exclusive** on-going chat memory. Using API with the same context array over and over is very useful. Without this feature, energy utilization increases and quality of response decreases. Often LLM come to different conclusions with the same input so that it sounds conversational. This is a huge problem trying to get back to the same state of understanding. Context parameter is critical. Please keep it.
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 feature 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#5665
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?