github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-17 05:36:16 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #15829] No Context Memory & Tool Calling Failure with Local Gemma4 via Ollama Docker #87776

New Issue
Open
opened 2026-05-10 06:22:40 -05:00 by GiteaMirror · 1 comment
GiteaMirror commented 2026-05-10 06:22:40 -05:00
Owner
Copy Link

Originally created by @OrzHex on GitHub (Apr 27, 2026).
Original GitHub issue: https://github.com/ollama/ollama/issues/15829

What is the issue?

I containerized Ollama using Docker and pulled the local gemma4:e2b / gemma4:26b models. I connect via port 11434. Both cc-haha (https://github.com/NanmiCoder/cc-haha ) and claude-code-tudou (https://github.com/AICoderTudou/claude-code-tudou ) can connect normally — the web interface loads and regular chat works.

However, two critical issues occur:
The conversation has no context memory (the model cannot remember previous messages in the chat).
Tool calling / function usage is not working.

How can I fix these problems?

Example of the context loss issue:
You: Hi, I am Tom. I am 8 years old.
AI: Hi Tom! Nice to meet you.
You: How old is Tom?
AI: I am gemma4. I do not have an age.

Example of the tool calling abnormal issue:
You: Open local file config.json.
AI: I can help you with file-related questions. Please tell me more details.

Relevant log output

OS

Docker

GPU

No response

CPU

Intel

Ollama version

0.21.1

Originally created by @OrzHex on GitHub (Apr 27, 2026). Original GitHub issue: https://github.com/ollama/ollama/issues/15829 ### What is the issue? I containerized Ollama using Docker and pulled the local gemma4:e2b / gemma4:26b models. I connect via port 11434. Both cc-haha (https://github.com/NanmiCoder/cc-haha ) and claude-code-tudou (https://github.com/AICoderTudou/claude-code-tudou ) can connect normally — the web interface loads and regular chat works. However, two critical issues occur: The conversation has no context memory (the model cannot remember previous messages in the chat). Tool calling / function usage is not working. How can I fix these problems? Example of the context loss issue: You: Hi, I am Tom. I am 8 years old. AI: Hi Tom! Nice to meet you. You: How old is Tom? AI: I am gemma4. I do not have an age. Example of the tool calling abnormal issue: You: Open local file config.json. AI: I can help you with file-related questions. Please tell me more details. ### Relevant log output ```shell ``` ### OS Docker ### GPU _No response_ ### CPU Intel ### Ollama version 0.21.1
GiteaMirror added the bug label 2026-05-10 06:22:40 -05:00
GiteaMirror commented 2026-05-10 06:22:43 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (May 8, 2026):

Context memory is maintained by the client.

$ docker run --rm -d --name ollama-15829 ollama/ollama:0.21.1
$ docker exec -it ollama-15829 ollama run gemma4:e2b --think=false
>>> Hi, I am Tom. I am 8 years old.
Hi Tom! It's very nice to meet you. I'm a large language model.

How are you doing today? Is there anything you'd like to talk about or anything you're curious about? 😊

>>> How old is Tom? 
You told me that **you** are 8 years old, Tom! So, according to what you told me, **Tom is 8 years old.** 😊

Tool failures when using an agent/harness are usually because the context is too small to hold the tool definitions. Increase the context length.

<!-- gh-comment-id:4406199576 --> @rick-github commented on GitHub (May 8, 2026): Context memory is maintained by the client. ```console $ docker run --rm -d --name ollama-15829 ollama/ollama:0.21.1 $ docker exec -it ollama-15829 ollama run gemma4:e2b --think=false >>> Hi, I am Tom. I am 8 years old. Hi Tom! It's very nice to meet you. I'm a large language model. How are you doing today? Is there anything you'd like to talk about or anything you're curious about? 😊 >>> How old is Tom? You told me that **you** are 8 years old, Tom! So, according to what you told me, **Tom is 8 years old.** 😊 ``` Tool failures when using an agent/harness are usually because the context is too small to hold the tool definitions. Increase the [context length](https://github.com/ollama/ollama/blob/main/docs/faq.mdx#how-can-i-specify-the-context-window-size).
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
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#87776
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?