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 #7461] Dual GPU token generation bug on 0.3.15 that did not exist on 0.3.13 #51254

New Issue
Closed
opened 2026-04-28 19:05:13 -05:00 by GiteaMirror · 1 comment
GiteaMirror commented 2026-04-28 19:05:13 -05:00
Owner
Copy Link

Originally created by @calmingaura on GitHub (Nov 1, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/7461

Originally assigned to: @dhiltgen on GitHub.

What is the issue?

I have two separate Windows 11 systems, each with dual Radeon 7900XTX 24GB VRAM GPUs for inferencing using 70b parameter models. When running Ollama using the command line interface on 0.3.13, both systems worked great. However, switching to 0.3.15 cause generation of gibberish on the CLI, with a lot of ///////////////*he//them//// ..... When I switched back to 0.3.13, the problem was fixed. This bug is repeatable on both systems which have slightly different configurations, so I'm pretty sure it is the update. When I run a small model and it loads only into one GPUs VRAM, there is no problem on 0.3.13 or 0.3.15. Thanks!

OS

Windows

GPU

AMD

CPU

Intel

Ollama version

0.3.13 and tried 0.3.15

Originally created by @calmingaura on GitHub (Nov 1, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/7461 Originally assigned to: @dhiltgen on GitHub. ### What is the issue? I have two separate Windows 11 systems, each with dual Radeon 7900XTX 24GB VRAM GPUs for inferencing using 70b parameter models. When running Ollama using the command line interface on 0.3.13, both systems worked great. However, switching to 0.3.15 cause generation of gibberish on the CLI, with a lot of ///////////////*he//them//// ..... When I switched back to 0.3.13, the problem was fixed. This bug is repeatable on both systems which have slightly different configurations, so I'm pretty sure it is the update. When I run a small model and it loads only into one GPUs VRAM, there is no problem on 0.3.13 or 0.3.15. Thanks! ### OS Windows ### GPU AMD ### CPU Intel ### Ollama version 0.3.13 and tried 0.3.15
GiteaMirror added the bugamdwindows labels 2026-04-28 19:05:14 -05:00
GiteaMirror commented 2026-04-28 19:05:19 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Nov 1, 2024):

Do you mean 0.3.14? Server logs from both versions might also help in debugging.

<!-- gh-comment-id:2451559809 --> @rick-github commented on GitHub (Nov 1, 2024): Do you mean 0.3.14? [Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) from both versions might also help in debugging.
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 amd bug windows
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#51254
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?