github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-16 05:05:37 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #8418] Running OLLAMA_FLASH_ATTENTION=true with LoRA Models Returns: flash_attn is not compatible with LoRA #5407

New Issue
Open
opened 2026-04-12 16:38:48 -05:00 by GiteaMirror · 0 comments
GiteaMirror commented 2026-04-12 16:38:48 -05:00
Owner
Copy Link

Originally created by @AXDIGI on GitHub (Jan 14, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/8418

What is the issue?

Hello, I use fine tuned LLMs that use LoRA, when activating OLLAMA_FLASH_ATTENTION=true ollama serve the fine tuned models do not work, the error received is:

llama_lora_adapter_set: flash_attn is not compatible with LoRA
panic: error applying lora from file

This error stops the model from running until you flag it false. Then it runs correctly when its turned to false.

I have found on the llama.cpp git that they edited and removed part of their code last week to fix this. Here was the same issue: https://github.com/ggerganov/llama.cpp/discussions/11097

Here is the fix that was added to llama.cpp:

https://github.com/ggerganov/llama.cpp/pull/11104/files

I'm not sure what I've done is correct in providing the above information (first ever report), but I found that Ollama uses Llama.cpp so I thought I'd find the solution directly.

OS

macOS

GPU

No response

CPU

No response

Ollama version

0.5.5

Originally created by @AXDIGI on GitHub (Jan 14, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/8418 ### What is the issue? Hello, I use fine tuned LLMs that use LoRA, when activating OLLAMA_FLASH_ATTENTION=true ollama serve the fine tuned models do not work, the error received is: ``` llama_lora_adapter_set: flash_attn is not compatible with LoRA panic: error applying lora from file ``` This error stops the model from running until you flag it false. Then it runs correctly when its turned to false. I have found on the llama.cpp git that they edited and removed part of their code last week to fix this. Here was the same issue: https://github.com/ggerganov/llama.cpp/discussions/11097 Here is the fix that was added to llama.cpp: https://github.com/ggerganov/llama.cpp/pull/11104/files I'm not sure what I've done is correct in providing the above information (first ever report), but I found that Ollama uses Llama.cpp so I thought I'd find the solution directly. ### OS macOS ### GPU _No response_ ### CPU _No response_ ### Ollama version 0.5.5
GiteaMirror added the bug label 2026-04-12 16:38:48 -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 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#5407
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?