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 #10730] Ollama ver. 0.7.0 force to update llama3.2-vision model (working in ver 0.6.0), but then it did not work #53560

New Issue
Closed
opened 2026-04-29 03:47:34 -05:00 by GiteaMirror · 10 comments
GiteaMirror commented 2026-04-29 03:47:34 -05:00
Owner
Copy Link

Originally created by @DanoPTT on GitHub (May 16, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/10730

What is the issue?

I updated to Ollama 0.7.0, than I try to run llama3.2-vision model (from ver 0.6). It refuses run, and ask to download actualized llama3.2-vision model. I do that, but now this model goes to error when trying to analyze picture.
Other models like gemma3:12b works still correctly.

Relevant log output

time=2025-05-16T13:11:57.887+02:00 level=INFO source=runner.go:836 msg="starting ollama engine"
time=2025-05-16T13:11:57.890+02:00 level=INFO source=runner.go:899 msg="Server listening on 127.0.0.1:60512"
time=2025-05-16T13:11:57.928+02:00 level=INFO source=ggml.go:73 msg="" architecture=mllama file_type=Q4_K_M name="" description="" num_tensors=908 num_key_values=39
load_backend: loaded CPU backend from C:\ollama\lib\ollama\ggml-cpu-haswell.dll
time=2025-05-16T13:11:57.949+02:00 level=WARN source=sched.go:676 msg="gpu VRAM usage didn't recover within timeout" seconds=5.2667797 runner.size="11.7 GiB" runner.vram="11.7 GiB" runner.parallel=1 runner.pid=26216 runner.model=C:\Users\petrik.MMSSOFTEC\.ollama\models\blobs\sha256-7633fdffe14c0f7acc115402376be5bd6052220c348676c5133dc011b35e2429
ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 1 CUDA devices:
  Device 0: NVIDIA GeForce RTX 5060 Ti, compute capability 12.0, VMM: yes
load_backend: loaded CUDA backend from C:\ollama\lib\ollama\cuda_v12\ggml-cuda.dll
time=2025-05-16T13:11:58.050+02:00 level=INFO source=ggml.go:104 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.BMI2=1 CPU.0.LLAMAFILE=1 CPU.1.LLAMAFILE=1 CUDA.0.ARCHS=500,600,610,700,750,800,860,870,890,900,1200 CUDA.0.USE_GRAPHS=1 CUDA.0.PEER_MAX_BATCH_SIZE=128 compiler=cgo(clang)
time=2025-05-16T13:11:58.092+02:00 level=INFO source=server.go:625 msg="waiting for server to become available" status="llm server loading model"
time=2025-05-16T13:11:58.167+02:00 level=INFO source=ggml.go:299 msg="model weights" buffer=CUDA0 size="7.2 GiB"
time=2025-05-16T13:11:58.167+02:00 level=INFO source=ggml.go:299 msg="model weights" buffer=CPU size="285.7 MiB"
time=2025-05-16T13:11:58.200+02:00 level=WARN source=sched.go:676 msg="gpu VRAM usage didn't recover within timeout" seconds=5.5183276 runner.size="11.7 GiB" runner.vram="11.7 GiB" runner.parallel=1 runner.pid=26216 runner.model=C:\Users\petrik.MMSSOFTEC\.ollama\models\blobs\sha256-7633fdffe14c0f7acc115402376be5bd6052220c348676c5133dc011b35e2429
time=2025-05-16T13:12:00.759+02:00 level=INFO source=ggml.go:556 msg="compute graph" backend=CUDA0 buffer_type=CUDA0 size="280.0 MiB"
time=2025-05-16T13:12:00.759+02:00 level=INFO source=ggml.go:556 msg="compute graph" backend=CPU buffer_type=CPU size="8.0 MiB"
time=2025-05-16T13:12:00.871+02:00 level=INFO source=server.go:630 msg="llama runner started in 3.04 seconds"
C:\a\ollama\ollama\ml\backend\ggml\ggml\src\ggml-cuda\pad.cu:44: GGML_ASSERT(src0->ne[3] == 1 && dst->ne[3] == 1) failed
[GIN] 2025/05/16 - 13:12:01 | 200 |    9.4290978s |       127.0.0.1 | POST     "/v1/chat/completions"
time=2025-05-16T13:12:02.044+02:00 level=ERROR source=server.go:457 msg="llama runner terminated" error="exit status 0xc0000409"

OS

Windows

GPU

Nvidia

CPU

AMD

Ollama version

0.7.0

Originally created by @DanoPTT on GitHub (May 16, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/10730 ### What is the issue? I updated to Ollama 0.7.0, than I try to run llama3.2-vision model (from ver 0.6). It refuses run, and ask to download actualized llama3.2-vision model. I do that, but now this model goes to error when trying to analyze picture. Other models like gemma3:12b works still correctly. ### Relevant log output ```shell time=2025-05-16T13:11:57.887+02:00 level=INFO source=runner.go:836 msg="starting ollama engine" time=2025-05-16T13:11:57.890+02:00 level=INFO source=runner.go:899 msg="Server listening on 127.0.0.1:60512" time=2025-05-16T13:11:57.928+02:00 level=INFO source=ggml.go:73 msg="" architecture=mllama file_type=Q4_K_M name="" description="" num_tensors=908 num_key_values=39 load_backend: loaded CPU backend from C:\ollama\lib\ollama\ggml-cpu-haswell.dll time=2025-05-16T13:11:57.949+02:00 level=WARN source=sched.go:676 msg="gpu VRAM usage didn't recover within timeout" seconds=5.2667797 runner.size="11.7 GiB" runner.vram="11.7 GiB" runner.parallel=1 runner.pid=26216 runner.model=C:\Users\petrik.MMSSOFTEC\.ollama\models\blobs\sha256-7633fdffe14c0f7acc115402376be5bd6052220c348676c5133dc011b35e2429 ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 CUDA devices: Device 0: NVIDIA GeForce RTX 5060 Ti, compute capability 12.0, VMM: yes load_backend: loaded CUDA backend from C:\ollama\lib\ollama\cuda_v12\ggml-cuda.dll time=2025-05-16T13:11:58.050+02:00 level=INFO source=ggml.go:104 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.BMI2=1 CPU.0.LLAMAFILE=1 CPU.1.LLAMAFILE=1 CUDA.0.ARCHS=500,600,610,700,750,800,860,870,890,900,1200 CUDA.0.USE_GRAPHS=1 CUDA.0.PEER_MAX_BATCH_SIZE=128 compiler=cgo(clang) time=2025-05-16T13:11:58.092+02:00 level=INFO source=server.go:625 msg="waiting for server to become available" status="llm server loading model" time=2025-05-16T13:11:58.167+02:00 level=INFO source=ggml.go:299 msg="model weights" buffer=CUDA0 size="7.2 GiB" time=2025-05-16T13:11:58.167+02:00 level=INFO source=ggml.go:299 msg="model weights" buffer=CPU size="285.7 MiB" time=2025-05-16T13:11:58.200+02:00 level=WARN source=sched.go:676 msg="gpu VRAM usage didn't recover within timeout" seconds=5.5183276 runner.size="11.7 GiB" runner.vram="11.7 GiB" runner.parallel=1 runner.pid=26216 runner.model=C:\Users\petrik.MMSSOFTEC\.ollama\models\blobs\sha256-7633fdffe14c0f7acc115402376be5bd6052220c348676c5133dc011b35e2429 time=2025-05-16T13:12:00.759+02:00 level=INFO source=ggml.go:556 msg="compute graph" backend=CUDA0 buffer_type=CUDA0 size="280.0 MiB" time=2025-05-16T13:12:00.759+02:00 level=INFO source=ggml.go:556 msg="compute graph" backend=CPU buffer_type=CPU size="8.0 MiB" time=2025-05-16T13:12:00.871+02:00 level=INFO source=server.go:630 msg="llama runner started in 3.04 seconds" C:\a\ollama\ollama\ml\backend\ggml\ggml\src\ggml-cuda\pad.cu:44: GGML_ASSERT(src0->ne[3] == 1 && dst->ne[3] == 1) failed [GIN] 2025/05/16 - 13:12:01 | 200 | 9.4290978s | 127.0.0.1 | POST "/v1/chat/completions" time=2025-05-16T13:12:02.044+02:00 level=ERROR source=server.go:457 msg="llama runner terminated" error="exit status 0xc0000409" ``` ### OS Windows ### GPU Nvidia ### CPU AMD ### Ollama version 0.7.0
GiteaMirror added the bug label 2026-04-29 03:47:34 -05:00
GiteaMirror commented 2026-04-29 03:47:37 -05:00
Author
Owner
Copy Link

@bastard9 commented on GitHub (May 19, 2025):

I have the same problem.

<!-- gh-comment-id:2889646950 --> @bastard9 commented on GitHub (May 19, 2025): I have the same problem.
GiteaMirror commented 2026-04-29 03:47:40 -05:00
Author
Owner
Copy Link

@junewinter2013 commented on GitHub (May 19, 2025):

We are having similar problem with llama3.2-vision on ollama 0.7.0 here also.
qwen2.5-vl is fine.

<!-- gh-comment-id:2890908224 --> @junewinter2013 commented on GitHub (May 19, 2025): We are having similar problem with llama3.2-vision on ollama 0.7.0 here also. qwen2.5-vl is fine.
GiteaMirror commented 2026-04-29 03:47:42 -05:00
Author
Owner
Copy Link

@HarrievG commented on GitHub (May 19, 2025):

same

GGML_ASSERT(sections.v[0] \u003e 0 || sections.v[1] \u003e 0 || sections.v[2] \u003e 0) failed

PM for payload

<!-- gh-comment-id:2891543251 --> @HarrievG commented on GitHub (May 19, 2025): same GGML_ASSERT(sections.v[0] \u003e 0 || sections.v[1] \u003e 0 || sections.v[2] \u003e 0) failed PM for payload
GiteaMirror commented 2026-04-29 03:47:43 -05:00
Author
Owner
Copy Link

@dontpanic121 commented on GitHub (May 21, 2025):

I encountered the issue where a model stopped running after an update. Unfortunately, I don't have a straightforward solution at the moment, except for attempting a roll-back to a previous version of Ollama

  • The problem started with v0.7.0.
  • The last working version was v0.6.8

The following message appears when testing Ollama in the CMD in Windows 11.

{"error":"'llama3.2-vision' is no longer compatible with your version of Ollama and has been replaced by a newer version. To re-download, run 'ollama pull llama3.2-vision'"}

It would be nice to know whether this issue is intended behavior, left unfixed by design, or something that needs to be addressed.

<!-- gh-comment-id:2896438828 --> @dontpanic121 commented on GitHub (May 21, 2025): I encountered the issue where a model stopped running after an update. Unfortunately, I don't have a straightforward solution at the moment, except for attempting a roll-back to a previous version of Ollama - The problem started with v0.7.0. - The last working version was v0.6.8 The following message appears when testing Ollama in the CMD in Windows 11. ```powershell {"error":"'llama3.2-vision' is no longer compatible with your version of Ollama and has been replaced by a newer version. To re-download, run 'ollama pull llama3.2-vision'"} ``` It would be nice to know whether this issue is intended behavior, left unfixed by design, or something that needs to be addressed.
GiteaMirror commented 2026-04-29 03:47:45 -05:00
Author
Owner
Copy Link

@DanoPTT commented on GitHub (May 21, 2025):

if you want to use your llama3.2.-vision model downgrade to 0.6.8. I upgraded to 0.7.0, then i got message like you, than download actualized model llama3.2.-vision, and since then I can not analyze pictures with that model. For text is model working, but for pictures not.

<!-- gh-comment-id:2896767500 --> @DanoPTT commented on GitHub (May 21, 2025): if you want to use your llama3.2.-vision model downgrade to 0.6.8. I upgraded to 0.7.0, then i got message like you, than download actualized model llama3.2.-vision, and since then I can not analyze pictures with that model. For text is model working, but for pictures not.
GiteaMirror commented 2026-04-29 03:47:46 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (May 24, 2025):

Model has been updated and 0.7.1 has a fix that should fix the crash.

<!-- gh-comment-id:2906093491 --> @rick-github commented on GitHub (May 24, 2025): Model has been updated and 0.7.1 has a fix that should fix the crash.
GiteaMirror commented 2026-04-29 03:47:47 -05:00
Author
Owner
Copy Link

@dontpanic121 commented on GitHub (May 24, 2025):

Model has been updated and 0.7.1 has a fix that should fix the crash.

I upgraded to 0.7.1 from 0.6.8 and the issue appears to still persist for me.

API Error: Error from Ollama: AI_RetryError: Failed after 3 attempts. Last error: 'llama3.2-vision' is no longer compatible with your version of Ollama and has been replaced by a newer version. To re-download, run 'ollama pull llama3.2-vision'
  • The last working version for me was v0.6.8
  • The problem started with v0.7.0
  • The problem still persist in v0.7.1
<!-- gh-comment-id:2906278143 --> @dontpanic121 commented on GitHub (May 24, 2025): > Model has been updated and 0.7.1 has a fix that should fix the crash. I upgraded to 0.7.1 from 0.6.8 and the issue appears to still persist for me. ```shell API Error: Error from Ollama: AI_RetryError: Failed after 3 attempts. Last error: 'llama3.2-vision' is no longer compatible with your version of Ollama and has been replaced by a newer version. To re-download, run 'ollama pull llama3.2-vision' ``` - The last working version for me was v0.6.8 - The problem started with v0.7.0 - The problem still persist in v0.7.1
GiteaMirror commented 2026-04-29 03:47:48 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (May 24, 2025):

Run the following on your ollama server:

ollama pull llama3.2-vision
<!-- gh-comment-id:2906509808 --> @rick-github commented on GitHub (May 24, 2025): Run the following on your ollama server: ``` ollama pull llama3.2-vision ```
GiteaMirror commented 2026-04-29 03:47:49 -05:00
Author
Owner
Copy Link

@dontpanic121 commented on GitHub (May 24, 2025):

Run the following on your ollama server:

ollama pull llama3.2-vision

Did not work and was forced to downgrade again to v0.6.8. I guess I am kind of stuck until something gets fixed or just live with it.

API Error: Error from Ollama: AI_RetryError: Failed after 3 attempts. Last error: 'llama3.2-vision' is no longer compatible with your version of Ollama and has been replaced by a newer version. To re-download, run 'ollama pull llama3.2-vision'
<!-- gh-comment-id:2907323100 --> @dontpanic121 commented on GitHub (May 24, 2025): > Run the following on your ollama server: > > ``` > ollama pull llama3.2-vision > ``` Did not work and was forced to downgrade again to v0.6.8. I guess I am kind of stuck until something gets fixed or just live with it. ```powershell API Error: Error from Ollama: AI_RetryError: Failed after 3 attempts. Last error: 'llama3.2-vision' is no longer compatible with your version of Ollama and has been replaced by a newer version. To re-download, run 'ollama pull llama3.2-vision' ```
GiteaMirror commented 2026-04-29 03:47:50 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (May 24, 2025):

Upgrade the ollama server and pull the model. It works for everybody else, if you still have a problem, logs will aid in debugging.

<!-- gh-comment-id:2907334780 --> @rick-github commented on GitHub (May 24, 2025): Upgrade the ollama server and pull the model. It works for everybody else, if you still have a problem, [logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) will aid 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 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#53560
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?