github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-17 13:47:08 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #11266] The model messed up memories #33187

New Issue
Closed
opened 2026-04-22 15:37:17 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-04-22 15:37:17 -05:00
Owner
Copy Link

Originally created by @UserEdmund on GitHub (Jul 2, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/11266

What is the issue?

When I run an ollama server with a model and then closed and then reopened it, the model keeps the memory of last responses and ignores my new questions.

Relevant log output

OS

No response

GPU

Apple

CPU

Apple

Ollama version

No response

Originally created by @UserEdmund on GitHub (Jul 2, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/11266 ### What is the issue? When I run an ollama server with a model and then closed and then reopened it, the model keeps the memory of last responses and ignores my new questions. ### Relevant log output ```shell ``` ### OS _No response_ ### GPU Apple ### CPU Apple ### Ollama version _No response_
GiteaMirror added the needs more infobug labels 2026-04-22 15:37:18 -05:00
GiteaMirror commented 2026-04-22 15:37:21 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Jul 2, 2025):

Can you give an example?

<!-- gh-comment-id:3027195344 --> @rick-github commented on GitHub (Jul 2, 2025): Can you give an example?
GiteaMirror commented 2026-04-22 15:37:22 -05:00
Author
Owner
Copy Link

@PlkMarudny commented on GitHub (Jul 8, 2025):

I have a kind of example: I use a simple Python script to send data to /generate endpoint. It happens to have an image attached:

payload = {
        "model": MODEL_NAME,
        "prompt": prompt_text,
        "images": [base64_image],
        "stream": False,
        "options": {
            "num_ctx": 12800
        }

Model is gemma3:12b. The prompt is around 2400 words long. The image is smaller Gemma allows. First request works. Issuing the same request the second time produces:

Okay, I'm ready to analyze the image. Please provide the Base64 encoded image data. I will then generate the JSON object as per the instructions.

To make it work, I have to issue another request, with a basic prompt, like "Describe the image", after that, my inital prompt works again. The same happens on the chat enpoint btw.

<!-- gh-comment-id:3048852845 --> @PlkMarudny commented on GitHub (Jul 8, 2025): I have a kind of example: I use a simple Python script to send data to /generate endpoint. It happens to have an image attached: ``` payload = { "model": MODEL_NAME, "prompt": prompt_text, "images": [base64_image], "stream": False, "options": { "num_ctx": 12800 } ``` Model is ```gemma3:12b```. The prompt is around 2400 words long. The image is smaller Gemma allows. First request works. Issuing the same request the second time produces: ``` Okay, I'm ready to analyze the image. Please provide the Base64 encoded image data. I will then generate the JSON object as per the instructions. ``` To make it work, I have to issue another request, with a basic prompt, like "Describe the image", after that, my inital prompt works again. The same happens on the ```chat``` enpoint btw.
GiteaMirror commented 2026-04-22 15:37:23 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 1, 2025):

If you could make a script that fully demonstrates the problem it will be easier to debug.

<!-- gh-comment-id:3243202331 --> @rick-github commented on GitHub (Sep 1, 2025): If you could make a script that fully demonstrates the problem it will be easier to debug.
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 bug needs more info
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#33187
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?