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 #9069] error while loading shared libraries: libggml_cuda_v11.so: ELF load command past end of file #5906

New Issue
Closed
opened 2026-04-12 17:14:34 -05:00 by GiteaMirror · 6 comments
GiteaMirror commented 2026-04-12 17:14:34 -05:00
Owner
Copy Link

Originally created by @xcsnx on GitHub (Feb 13, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/9069

What is the issue?

environment：

Image

Image

Image

ollama-linux-amd64.tgz

bug：

Image
/opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_llama_server: error while loading shared libraries: libggml_cuda_v11.so: ELF load command past end o

Image

How should this error be handled？

Relevant log output

Feb 13 20:50:48 aiops003 ollama[1214]: time=2025-02-13T20:50:48.943+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="/opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_
Feb 13 20:50:48 aiops003 ollama[1214]: time=2025-02-13T20:50:48.943+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1
Feb 13 20:50:48 aiops003 ollama[1214]: time=2025-02-13T20:50:48.943+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding"
Feb 13 20:50:48 aiops003 ollama[1214]: /opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_llama_server: error while loading shared libraries: libggml_cuda_v11.so: ELF load command past end o
Feb 13 20:50:48 aiops003 ollama[1214]: time=2025-02-13T20:50:48.944+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error"
Feb 13 20:50:49 aiops003 ollama[1214]: time=2025-02-13T20:50:49.194+08:00 level=ERROR source=sched.go:455 msg="error loading llama server" error="llama runner process has terminated: exit st
Feb 13 20:50:49 aiops003 ollama[1214]: [GIN] 2025/02/13 - 20:50:49 | 500 |  634.140215ms |       127.0.0.1 | POST     "/api/generate"
Feb 13 20:50:54 aiops003 ollama[1214]: time=2025-02-13T20:50:54.336+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.142091008 model=/opt/oll
Feb 13 20:50:54 aiops003 ollama[1214]: time=2025-02-13T20:50:54.587+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.392626392 model=/opt/oll
Feb 13 20:50:54 aiops003 ollama[1214]: time=2025-02-13T20:50:54.837+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.642826874 model=/opt/oll
Feb 13 21:14:52 aiops003 ollama[1214]: [GIN] 2025/02/13 - 21:14:52 | 200 |      69.284µs |       127.0.0.1 | HEAD     "/"
Feb 13 21:14:52 aiops003 ollama[1214]: [GIN] 2025/02/13 - 21:14:52 | 200 |    31.41354ms |       127.0.0.1 | POST     "/api/show"
Feb 13 21:14:52 aiops003 ollama[1214]: time=2025-02-13T21:14:52.942+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=/opt/ollama/m
Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.090+08:00 level=INFO source=server.go:104 msg="system memory" total="62.2 GiB" free="29.9 GiB" free_swap="31.2 GiB"
Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.091+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=29 layers.offload=29 layers.s
Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.092+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="/opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_
Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.092+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1
Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.092+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding"
Feb 13 21:14:53 aiops003 ollama[1214]: /opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_llama_server: error while loading shared libraries: libggml_cuda_v11.so: ELF load command past end o
Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.092+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error"
Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.343+08:00 level=ERROR source=sched.go:455 msg="error loading llama server" error="llama runner process has terminated: exit st
Feb 13 21:14:53 aiops003 ollama[1214]: [GIN] 2025/02/13 - 21:14:53 | 500 |   653.90216ms |       127.0.0.1 | POST     "/api/generate"
Feb 13 21:14:54 aiops003 ollama[1214]: [GIN] 2025/02/13 - 21:14:54 | 200 |      50.188µs |       127.0.0.1 | HEAD     "/"
Feb 13 21:14:54 aiops003 ollama[1214]: [GIN] 2025/02/13 - 21:14:54 | 200 |   33.526529ms |       127.0.0.1 | POST     "/api/show"
Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.132+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=/opt/ollama/m
Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.310+08:00 level=INFO source=server.go:104 msg="system memory" total="62.2 GiB" free="29.9 GiB" free_swap="31.2 GiB"
Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.310+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=29 layers.offload=29 layers.s
Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.311+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="/opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_
Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.311+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1
Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.311+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding"
Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.312+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error"
Feb 13 21:14:55 aiops003 ollama[1214]: /opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_llama_server: error while loading shared libraries: libggml_cuda_v11.so: ELF load command past end o
Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.562+08:00 level=ERROR source=sched.go:455 msg="error loading llama server" error="llama runner process has terminated: exit st
Feb 13 21:14:55 aiops003 ollama[1214]: [GIN] 2025/02/13 - 21:14:55 | 500 |  645.989673ms |       127.0.0.1 | POST     "/api/generate"
Feb 13 21:14:58 aiops003 ollama[1214]: time=2025-02-13T21:14:58.491+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.147951666 model=/opt/oll
Feb 13 21:14:58 aiops003 ollama[1214]: time=2025-02-13T21:14:58.741+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.397326342 model=/opt/oll
Feb 13 21:14:58 aiops003 ollama[1214]: time=2025-02-13T21:14:58.990+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.647177827 model=/opt/oll

OS

Linux

GPU

Nvidia

CPU

Intel

Ollama version

ollama version is 0.5.7

Originally created by @xcsnx on GitHub (Feb 13, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/9069 ### What is the issue? environment： ![Image](https://github.com/user-attachments/assets/6a6b10e2-3cf3-4ab9-86b1-face70ac96aa) ![Image](https://github.com/user-attachments/assets/ca3bd6b4-53d5-431d-9b01-9f1e5bef276d) ![Image](https://github.com/user-attachments/assets/1fa6ffd6-002d-4745-89b8-8667d958169c) ollama-linux-amd64.tgz bug： ![Image](https://github.com/user-attachments/assets/7e8217bf-0e9e-4f5b-8407-2f2fe229dec2) /opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_llama_server: error while loading shared libraries: libggml_cuda_v11.so: ELF load command past end o ![Image](https://github.com/user-attachments/assets/e8042386-a16b-45bd-b91d-72370ea2370d) How should this error be handled？ ### Relevant log output ```shell Feb 13 20:50:48 aiops003 ollama[1214]: time=2025-02-13T20:50:48.943+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="/opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_ Feb 13 20:50:48 aiops003 ollama[1214]: time=2025-02-13T20:50:48.943+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1 Feb 13 20:50:48 aiops003 ollama[1214]: time=2025-02-13T20:50:48.943+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding" Feb 13 20:50:48 aiops003 ollama[1214]: /opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_llama_server: error while loading shared libraries: libggml_cuda_v11.so: ELF load command past end o Feb 13 20:50:48 aiops003 ollama[1214]: time=2025-02-13T20:50:48.944+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error" Feb 13 20:50:49 aiops003 ollama[1214]: time=2025-02-13T20:50:49.194+08:00 level=ERROR source=sched.go:455 msg="error loading llama server" error="llama runner process has terminated: exit st Feb 13 20:50:49 aiops003 ollama[1214]: [GIN] 2025/02/13 - 20:50:49 | 500 | 634.140215ms | 127.0.0.1 | POST "/api/generate" Feb 13 20:50:54 aiops003 ollama[1214]: time=2025-02-13T20:50:54.336+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.142091008 model=/opt/oll Feb 13 20:50:54 aiops003 ollama[1214]: time=2025-02-13T20:50:54.587+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.392626392 model=/opt/oll Feb 13 20:50:54 aiops003 ollama[1214]: time=2025-02-13T20:50:54.837+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.642826874 model=/opt/oll Feb 13 21:14:52 aiops003 ollama[1214]: [GIN] 2025/02/13 - 21:14:52 | 200 | 69.284µs | 127.0.0.1 | HEAD "/" Feb 13 21:14:52 aiops003 ollama[1214]: [GIN] 2025/02/13 - 21:14:52 | 200 | 31.41354ms | 127.0.0.1 | POST "/api/show" Feb 13 21:14:52 aiops003 ollama[1214]: time=2025-02-13T21:14:52.942+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=/opt/ollama/m Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.090+08:00 level=INFO source=server.go:104 msg="system memory" total="62.2 GiB" free="29.9 GiB" free_swap="31.2 GiB" Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.091+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=29 layers.offload=29 layers.s Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.092+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="/opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_ Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.092+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1 Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.092+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding" Feb 13 21:14:53 aiops003 ollama[1214]: /opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_llama_server: error while loading shared libraries: libggml_cuda_v11.so: ELF load command past end o Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.092+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error" Feb 13 21:14:53 aiops003 ollama[1214]: time=2025-02-13T21:14:53.343+08:00 level=ERROR source=sched.go:455 msg="error loading llama server" error="llama runner process has terminated: exit st Feb 13 21:14:53 aiops003 ollama[1214]: [GIN] 2025/02/13 - 21:14:53 | 500 | 653.90216ms | 127.0.0.1 | POST "/api/generate" Feb 13 21:14:54 aiops003 ollama[1214]: [GIN] 2025/02/13 - 21:14:54 | 200 | 50.188µs | 127.0.0.1 | HEAD "/" Feb 13 21:14:54 aiops003 ollama[1214]: [GIN] 2025/02/13 - 21:14:54 | 200 | 33.526529ms | 127.0.0.1 | POST "/api/show" Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.132+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=/opt/ollama/m Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.310+08:00 level=INFO source=server.go:104 msg="system memory" total="62.2 GiB" free="29.9 GiB" free_swap="31.2 GiB" Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.310+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=29 layers.offload=29 layers.s Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.311+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="/opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_ Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.311+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1 Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.311+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding" Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.312+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error" Feb 13 21:14:55 aiops003 ollama[1214]: /opt/ollama/lib/ollama/runners/cuda_v11_avx/ollama_llama_server: error while loading shared libraries: libggml_cuda_v11.so: ELF load command past end o Feb 13 21:14:55 aiops003 ollama[1214]: time=2025-02-13T21:14:55.562+08:00 level=ERROR source=sched.go:455 msg="error loading llama server" error="llama runner process has terminated: exit st Feb 13 21:14:55 aiops003 ollama[1214]: [GIN] 2025/02/13 - 21:14:55 | 500 | 645.989673ms | 127.0.0.1 | POST "/api/generate" Feb 13 21:14:58 aiops003 ollama[1214]: time=2025-02-13T21:14:58.491+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.147951666 model=/opt/oll Feb 13 21:14:58 aiops003 ollama[1214]: time=2025-02-13T21:14:58.741+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.397326342 model=/opt/oll Feb 13 21:14:58 aiops003 ollama[1214]: time=2025-02-13T21:14:58.990+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.647177827 model=/opt/oll ``` ### OS Linux ### GPU Nvidia ### CPU Intel ### Ollama version ollama version is 0.5.7
GiteaMirror added the bug label 2026-04-12 17:14:34 -05:00
GiteaMirror commented 2026-04-12 17:14:36 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 13, 2025):

Did you run out of disk space while installing? Your libggml_cuda_v11.so is shorter, has the wrong mode, and hasn't had the mtime updated.

$ ls -l
total 965800
-rwxr-xr-x 1 root root 979085896 Jan 16 18:01 libggml_cuda_v11.so
-rwxr-xr-x 1 root root   9885296 Jan 16 18:03 ollama_llama_server
$ file libggml_cuda_v11.so
libggml_cuda_v11.so: ELF 64-bit LSB shared object, x86-64, version 1 (GNU/Linux), dynamically linked, BuildID[sha1]=de5f4b5eb1f5d6a71df36ad861957c886c945230, not stripped
<!-- gh-comment-id:2656736598 --> @rick-github commented on GitHub (Feb 13, 2025): Did you run out of disk space while installing? Your libggml_cuda_v11.so is shorter, has the wrong mode, and hasn't had the mtime updated. ```console $ ls -l total 965800 -rwxr-xr-x 1 root root 979085896 Jan 16 18:01 libggml_cuda_v11.so -rwxr-xr-x 1 root root 9885296 Jan 16 18:03 ollama_llama_server $ file libggml_cuda_v11.so libggml_cuda_v11.so: ELF 64-bit LSB shared object, x86-64, version 1 (GNU/Linux), dynamically linked, BuildID[sha1]=de5f4b5eb1f5d6a71df36ad861957c886c945230, not stripped
GiteaMirror commented 2026-04-12 17:14:36 -05:00
Author
Owner
Copy Link

@xcsnx commented on GitHub (Feb 13, 2025):

Did you run out of disk space while installing? Your libggml_cuda_v11.so is shorter, has the wrong mode, and hasn't had the mtime updated.

$ ls -l
total 965800
-rwxr-xr-x 1 root root 979085896 Jan 16 18:01 libggml_cuda_v11.so
-rwxr-xr-x 1 root root 9885296 Jan 16 18:03 ollama_llama_server
$ file libggml_cuda_v11.so
libggml_cuda_v11.so: ELF 64-bit LSB shared object, x86-64, version 1 (GNU/Linux), dynamically linked, BuildID[sha1]=de5f4b5eb1f5d6a71df36ad861957c886c945230, not stripped

Thank you for your reply. My disk space is sufficient, but I don’t quite understand what you mean. I installed it manually offline using the command tar -xzf ollama-linux-amd64.tgz. For details, please refer to: https://blog.csdn.net/weixin_44988127/article/details/145021587. Could you please provide further clarification?

<!-- gh-comment-id:2656797712 --> @xcsnx commented on GitHub (Feb 13, 2025): > Did you run out of disk space while installing? Your libggml_cuda_v11.so is shorter, has the wrong mode, and hasn't had the mtime updated. > > $ ls -l > total 965800 > -rwxr-xr-x 1 root root 979085896 Jan 16 18:01 libggml_cuda_v11.so > -rwxr-xr-x 1 root root 9885296 Jan 16 18:03 ollama_llama_server > $ file libggml_cuda_v11.so > libggml_cuda_v11.so: ELF 64-bit LSB shared object, x86-64, version 1 (GNU/Linux), dynamically linked, BuildID[sha1]=de5f4b5eb1f5d6a71df36ad861957c886c945230, not stripped Thank you for your reply. My disk space is sufficient, but I don’t quite understand what you mean. I installed it manually offline using the command tar -xzf ollama-linux-amd64.tgz. For details, please refer to: https://blog.csdn.net/weixin_44988127/article/details/145021587. Could you please provide further clarification?
GiteaMirror commented 2026-04-12 17:14:36 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 13, 2025):

Your installation is broken.

Do this:

tar zvtf ollama-linux-amd64.tgz ./lib/ollama/runners/cuda_v11_avx

Do the results look like this:

drwxr-xr-x root/root         0 2025-01-16 18:03 ./lib/ollama/runners/cuda_v11_avx/
-rwxr-xr-x root/root   9885296 2025-01-16 18:03 ./lib/ollama/runners/cuda_v11_avx/ollama_llama_server
-rwxr-xr-x root/root 979085896 2025-01-16 18:01 ./lib/ollama/runners/cuda_v11_avx/libggml_cuda_v11.so

Then compare with the files in /opt/ollama/lib/ollama/runners/cuda_v11_avx/:

ls -l /opt/ollama/lib/ollama/runners/cuda_v11_avx/

Do they match the output of tar?

Recommendation: re-install.

<!-- gh-comment-id:2656849207 --> @rick-github commented on GitHub (Feb 13, 2025): Your installation is broken. Do this: ``` tar zvtf ollama-linux-amd64.tgz ./lib/ollama/runners/cuda_v11_avx ``` Do the results look like this: ``` drwxr-xr-x root/root 0 2025-01-16 18:03 ./lib/ollama/runners/cuda_v11_avx/ -rwxr-xr-x root/root 9885296 2025-01-16 18:03 ./lib/ollama/runners/cuda_v11_avx/ollama_llama_server -rwxr-xr-x root/root 979085896 2025-01-16 18:01 ./lib/ollama/runners/cuda_v11_avx/libggml_cuda_v11.so ``` Then compare with the files in `/opt/ollama/lib/ollama/runners/cuda_v11_avx/`: ``` ls -l /opt/ollama/lib/ollama/runners/cuda_v11_avx/ ``` Do they match the output of `tar`? Recommendation: re-install.
GiteaMirror commented 2026-04-12 17:14:37 -05:00
Author
Owner
Copy Link

@xcsnx commented on GitHub (Feb 13, 2025):

Your installation is broken.

Do this:

tar zvtf ollama-linux-amd64.tgz ./lib/ollama/runners/cuda_v11_avx

Do the results look like this:

drwxr-xr-x root/root         0 2025-01-16 18:03 ./lib/ollama/runners/cuda_v11_avx/
-rwxr-xr-x root/root   9885296 2025-01-16 18:03 ./lib/ollama/runners/cuda_v11_avx/ollama_llama_server
-rwxr-xr-x root/root 979085896 2025-01-16 18:01 ./lib/ollama/runners/cuda_v11_avx/libggml_cuda_v11.so

Then compare with the files in /opt/ollama/lib/ollama/runners/cuda_v11_avx/:

ls -l /opt/ollama/lib/ollama/runners/cuda_v11_avx/

Do they match the output of tar?

Recommendation: re-install.

[root@aiops003 ollama]# tar zvtf ollama-linux-amd64.tgz ./lib/ollama/runners/cuda_v11_avx
drwxr-xr-x root/root 0 2025-01-17 01:03 ./lib/ollama/runners/cuda_v11_avx/
-rwxr-xr-x root/root 9885296 2025-01-17 01:03 ./lib/ollama/runners/cuda_v11_avx/ollama_llama_server
-rwxr-xr-x root/root 979085896 2025-01-17 01:01 ./lib/ollama/runners/cuda_v11_avx/libggml_cuda_v11.so

Thank you! As you said, the length of libggml_cuda_v11.so is indeed different from the one in the tar package! So can I directly replace the libggml_cuda_v11.so file?

<!-- gh-comment-id:2656882928 --> @xcsnx commented on GitHub (Feb 13, 2025): > Your installation is broken. > > Do this: > > ``` > tar zvtf ollama-linux-amd64.tgz ./lib/ollama/runners/cuda_v11_avx > ``` > > Do the results look like this: > > ``` > drwxr-xr-x root/root 0 2025-01-16 18:03 ./lib/ollama/runners/cuda_v11_avx/ > -rwxr-xr-x root/root 9885296 2025-01-16 18:03 ./lib/ollama/runners/cuda_v11_avx/ollama_llama_server > -rwxr-xr-x root/root 979085896 2025-01-16 18:01 ./lib/ollama/runners/cuda_v11_avx/libggml_cuda_v11.so > ``` > > Then compare with the files in `/opt/ollama/lib/ollama/runners/cuda_v11_avx/`: > > ``` > ls -l /opt/ollama/lib/ollama/runners/cuda_v11_avx/ > ``` > > Do they match the output of `tar`? > > Recommendation: re-install. [root@aiops003 ollama]# tar zvtf ollama-linux-amd64.tgz ./lib/ollama/runners/cuda_v11_avx drwxr-xr-x root/root 0 2025-01-17 01:03 ./lib/ollama/runners/cuda_v11_avx/ -rwxr-xr-x root/root 9885296 2025-01-17 01:03 ./lib/ollama/runners/cuda_v11_avx/ollama_llama_server -rwxr-xr-x root/root 979085896 2025-01-17 01:01 ./lib/ollama/runners/cuda_v11_avx/libggml_cuda_v11.so Thank you! As you said, the length of libggml_cuda_v11.so is indeed different from the one in the tar package! So can I directly replace the libggml_cuda_v11.so file?
GiteaMirror commented 2026-04-12 17:14:37 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 13, 2025):

There may be other incomplete files. Just re-install.

sudo tar --extract --verbose --file ollama-linux-amd64.tgz --directory /opt/ollama
<!-- gh-comment-id:2656916174 --> @rick-github commented on GitHub (Feb 13, 2025): There may be other incomplete files. Just re-install. ``` sudo tar --extract --verbose --file ollama-linux-amd64.tgz --directory /opt/ollama ```
GiteaMirror commented 2026-04-12 17:14:38 -05:00
Author
Owner
Copy Link

@xcsnx commented on GitHub (Feb 13, 2025):

Your installation is broken.

Do this:

tar zvtf ollama-linux-amd64.tgz ./lib/ollama/runners/cuda_v11_avx

Do the results look like this:

drwxr-xr-x root/root         0 2025-01-16 18:03 ./lib/ollama/runners/cuda_v11_avx/
-rwxr-xr-x root/root   9885296 2025-01-16 18:03 ./lib/ollama/runners/cuda_v11_avx/ollama_llama_server
-rwxr-xr-x root/root 979085896 2025-01-16 18:01 ./lib/ollama/runners/cuda_v11_avx/libggml_cuda_v11.so

Then compare with the files in /opt/ollama/lib/ollama/runners/cuda_v11_avx/:

ls -l /opt/ollama/lib/ollama/runners/cuda_v11_avx/

Do they match the output of tar?

Recommendation: re-install.

There may be other incomplete files. Just re-install.

sudo tar --extract --verbose --file ollama-linux-amd64.tgz --directory /opt/ollama

I directly replaced the libggml_cuda_v11.so file, and Ollama is now running on the GPU. Thank you!

<!-- gh-comment-id:2656924234 --> @xcsnx commented on GitHub (Feb 13, 2025): > Your installation is broken. > > Do this: > > ``` > tar zvtf ollama-linux-amd64.tgz ./lib/ollama/runners/cuda_v11_avx > ``` > > Do the results look like this: > > ``` > drwxr-xr-x root/root 0 2025-01-16 18:03 ./lib/ollama/runners/cuda_v11_avx/ > -rwxr-xr-x root/root 9885296 2025-01-16 18:03 ./lib/ollama/runners/cuda_v11_avx/ollama_llama_server > -rwxr-xr-x root/root 979085896 2025-01-16 18:01 ./lib/ollama/runners/cuda_v11_avx/libggml_cuda_v11.so > ``` > > Then compare with the files in `/opt/ollama/lib/ollama/runners/cuda_v11_avx/`: > > ``` > ls -l /opt/ollama/lib/ollama/runners/cuda_v11_avx/ > ``` > > Do they match the output of `tar`? > > Recommendation: re-install. > There may be other incomplete files. Just re-install. > > ``` > sudo tar --extract --verbose --file ollama-linux-amd64.tgz --directory /opt/ollama > ``` I directly replaced the libggml_cuda_v11.so file, and Ollama is now running on the GPU. Thank you!
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#5906
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?