ollama serve启动服务时检测不到显卡驱动

最新推荐文章于 2025-02-25 22:05:35 发布

晚安，chris

最新推荐文章于 2025-02-25 22:05:35 发布

阅读量3.9k

点赞数 4

分类专栏： rag应用开发文章标签：深度学习人工智能 ubuntu

本文链接：https://blog.csdn.net/qq_39918087/article/details/144530107

版权

rag应用开发专栏收录该内容

1 篇文章

订阅专栏

在学习rag应用开发时遇到ollama 启动服务时GPU检测不到的问题：

OLLAMA_HOST=:11436 ollama serve

终端打印信息：

2024/12/16 23:34:01 routes.go:1195: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY:http://127.0.0.1:8889/ HTTP_PROXY:http://127.0.0.1:8889/ NO_PROXY:localhost,127.0.0.0/8,::1 OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://:11436 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/home/chris/.ollama/models OLLAMA_MULTIUSER_CACHE:false OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:0 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://*] OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES: http_proxy:http://127.0.0.1:8889/ https_proxy:http://127.0.0.1:8889/ no_proxy:localhost,127.0.0.0/8,::1]"
time=2024-12-16T23:34:01.882+08:00 level=INFO source=images.go:753 msg="total blobs: 0"
time=2024-12-16T23:34:01.882+08:00 level=INFO source=images.go:760 msg="total unused blobs removed: 0"
time=2024-12-16T23:34:01.883+08:00 level=INFO source=routes.go:1246 msg="Listening on [::]:11436 (version 0.5.1)"
time=2024-12-16T23:34:01.883+08:00 level=INFO source=common.go:135 msg="extracting embedded files" dir=/tmp/ollama158827916/runners
time=2024-12-16T23:34:02.039+08:00 level=INFO source=common.go:49 msg="Dynamic LLM libraries" runners="[cuda_v12 rocm cpu cpu_avx cpu_avx2 cuda_v11]"
time=2024-12-16T23:34:02.039+08:00 level=INFO source=gpu.go:221 msg="looking for compatible GPUs"
time=2024-12-16T23:34:02.088+08:00 level=WARN source=gpu.go:613 msg="unknown error initializing cuda driver library /usr/lib/x86_64-linux-gnu/libcuda.so.535.216.01: cuda driver library init failure: 999. see https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md for more information"
time=2024-12-16T23:34:02.105+08:00 level=INFO source=gpu.go:386 msg="no compatible GPUs were discovered"
time=2024-12-16T23:34:02.105+08:00 level=INFO source=types.go:123 msg="inference compute" id=0 library=cpu variant=avx2 compute="" driver=0.0 name="" total="15.5 GiB" available="10.6 GiB"

其中的一条检测信息显示没有检测到显卡驱动：

time=2024-12-16T23:34:02.088+08:00 level=WARN source=gpu.go:613 msg="unknown error initializing cuda driver library /usr/lib/x86_64-linux-gnu/libcuda.so.535.216.01: cuda driver library init failure: 999. see https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md for more information"

查看官网报错日志后发现，是显卡驱动加载失败；只需要重新加载就好：

#先卸载
sudo rmmod nvidia_uvm
#再加载
sudo modprobe nvidia_uvm

重新启动服务后显卡驱动就可以检测到显卡驱动了

time=2024-12-17T10:51:17.621+08:00 level=INFO source=types.go:123 msg="inference compute" id=GPU-0ebdcbff-beee-7b2a-dbdf-702b95bac601 library=cuda variant=v12 compute=6.1 driver=12.2 name="NVIDIA GeForce GTX 1050 Ti" total="3.9 GiB" available="3.9 GiB"