github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 00:22:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #4673] BUG: PHI-3 #2938

New Issue
Open
opened 2026-04-12 13:18:35 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-04-12 13:18:35 -05:00
Owner
Copy Link

Originally created by @MichaelFomenko on GitHub (May 28, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/4673

What is the issue?

When I start the Conversation in German Language, Phi-3 Mini and Medium working fine. But after some Conversations, the Models starting producing slowly Gibberish and Nonsens and repeating phrases, word and tokens and don't answering my Quaestiones anymore. When I start a new Conversation, it works fine.

Open Web UI Version: 1.1.125

OS

Linux

GPU

AMD

CPU

AMD

Ollama version

0.1.39

Originally created by @MichaelFomenko on GitHub (May 28, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/4673 ### What is the issue? When I start the Conversation in German Language, Phi-3 Mini and Medium working fine. But after some Conversations, the Models starting producing slowly Gibberish and Nonsens and repeating phrases, word and tokens and don't answering my Quaestiones anymore. When I start a new Conversation, it works fine. Open Web UI Version: 1.1.125 ### OS Linux ### GPU AMD ### CPU AMD ### Ollama version 0.1.39
GiteaMirror added the bug label 2026-04-12 13:18:35 -05:00
GiteaMirror commented 2026-04-12 13:18:37 -05:00
Author
Owner
Copy Link

@jaasperhelin commented on GitHub (May 29, 2024):

I have the same issue, especially with RAG. Even after upgrading to the newest version of ollama today. Issue appears both with Phi 3 Mini and Medium, context window is set to 4000 tokens.
I use the prompt templates suggested by ollama. Does it require another template for RAG?

<!-- gh-comment-id:2137297809 --> @jaasperhelin commented on GitHub (May 29, 2024): I have the same issue, especially with RAG. Even after upgrading to the newest version of ollama today. Issue appears both with Phi 3 Mini and Medium, context window is set to 4000 tokens. I use the prompt templates suggested by ollama. Does it require another template for RAG?
GiteaMirror commented 2026-04-12 13:18:38 -05:00
Author
Owner
Copy Link

@MichaelFomenko commented on GitHub (May 30, 2024):

is here someone, who cares about BUGs? Maybee the Developer should not focus on new Features and make the existing code stable. And only implement new features whet it is absolutely necessary. Because every new feature needs testing time and with time the complexity if the Project will increase and if the base code is not testet the whole project fail, because of the BUGs.

<!-- gh-comment-id:2139185187 --> @MichaelFomenko commented on GitHub (May 30, 2024): is here someone, who cares about BUGs? Maybee the Developer should not focus on new Features and make the existing code stable. And only implement new features whet it is absolutely necessary. Because every new feature needs testing time and with time the complexity if the Project will increase and if the base code is not testet the whole project fail, because of the BUGs.
GiteaMirror commented 2026-04-12 13:18:39 -05:00
Author
Owner
Copy Link

@nils-se commented on GitHub (Jun 25, 2024):

Having the same issue. phi3 3B was very fast on my 4 core cpu. now its painfully slow. What changed is the ollama version and the phi3 version. I can't get the old version back, as I deleted it before the update, but its probably 10x slower. Gemma still gives solid performance.
OS: Kubuntu 22.04
ollama version: 0.1.45
CPU: Intel
no GPU

<!-- gh-comment-id:2188254238 --> @nils-se commented on GitHub (Jun 25, 2024): Having the same issue. phi3 3B was very fast on my 4 core cpu. now its painfully slow. What changed is the ollama version and the phi3 version. I can't get the old version back, as I deleted it before the update, but its probably 10x slower. Gemma still gives solid performance. OS: Kubuntu 22.04 ollama version: 0.1.45 CPU: Intel no GPU
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#2938
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?