Duix.Heygem
Duix.Heygem copied to clipboard
duixcom
heygem-gen-video启动时提示找不到显卡驱动
你好，如题。我已正确安装nvidia显卡和驱动，但依然其实找不到驱动，请问有无办法解决？谢谢。 heygem-gen-video反复重新启动，log提示如下。
2025-04-15 21:01:15 Traceback (most recent call last):
2025-04-15 21:01:15 File "/code/app_local.py", line 231, in <module>
2025-04-15 21:01:15 TransDhTask.instance()
2025-04-15 21:01:15 File "trans_dh_service.py", line 1207, in trans_dh_service.TransDhTask.instance
2025-04-15 21:01:15 File "trans_dh_service.py", line 1189, in trans_dh_service.TransDhTask.__init__
2025-04-15 21:01:15 File "compute_ctc_att_bnf.py", line 130, in compute_ctc_att_bnf.load_ppg_model
2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1152, in to
2025-04-15 21:01:15 return self._apply(convert)
2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 802, in _apply
2025-04-15 21:01:15 module._apply(fn)
2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 802, in _apply
2025-04-15 21:01:15 module._apply(fn)
2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 802, in _apply
2025-04-15 21:01:15 module._apply(fn)
2025-04-15 21:01:15 [Previous line repeated 1 more time]
2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 825, in _apply
2025-04-15 21:01:15 param_applied = fn(param)
2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1150, in convert
2025-04-15 21:01:15 return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None, non_blocking)
2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/cuda/__init__.py", line 302, in _lazy_init
2025-04-15 21:01:15 torch._C._cuda_init()
2025-04-15 21:01:15 RuntimeError: Found no NVIDIA driver on your system. Please check that you have an NVIDIA GPU and installed a driver from http://www.nvidia.com/Download/index.aspx
2025-04-15 21:01:18 2025-04-15 21:01:18,098 - cv2box - INFO - Use default multi mode: multi-thread, or you can set env 'CV_MULTI_MODE' to multi-process/torch-process
C:\>nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2024 NVIDIA Corporation
Built on Thu_Sep_12_02:55:00_Pacific_Daylight_Time_2024
Cuda compilation tools, release 12.6, V12.6.77
Build cuda_12.6.r12.6/compiler.34841621_0
C:\>nvidia-smi
Tue Apr 15 20:51:35 2025
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 572.83 Driver Version: 572.83 CUDA Version: 12.8 |
|-----------------------------------------+------------------------+----------------------+
| GPU Name Driver-Model | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+========================+======================|
| 0 NVIDIA GeForce RTX 2080 Ti WDDM | 00000000:01:00.0 On | N/A |
| 27% 35C P8 14W / 250W | 1183MiB / 22528MiB | 3% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+
+-----------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
已确认过所有的Issues。 已使用最新的docker-compose.yml windows：windows11 24H2 Docker desktop：4.35.1 Nodejs：18.20.8 WSL： 2.4.13
你好，如题。我已正确安装nvidia显卡和驱动，但依然其实找不到驱动，请问有无办法解决？谢谢。 heygem-gen-video反复重新启动，log提示如下。
2025-04-15 21:01:15 Traceback (most recent call last): 2025-04-15 21:01:15 File "/code/app_local.py", line 231, in <module> 2025-04-15 21:01:15 TransDhTask.instance() 2025-04-15 21:01:15 File "trans_dh_service.py", line 1207, in trans_dh_service.TransDhTask.instance 2025-04-15 21:01:15 File "trans_dh_service.py", line 1189, in trans_dh_service.TransDhTask.__init__ 2025-04-15 21:01:15 File "compute_ctc_att_bnf.py", line 130, in compute_ctc_att_bnf.load_ppg_model 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1152, in to 2025-04-15 21:01:15 return self._apply(convert) 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 802, in _apply 2025-04-15 21:01:15 module._apply(fn) 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 802, in _apply 2025-04-15 21:01:15 module._apply(fn) 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 802, in _apply 2025-04-15 21:01:15 module._apply(fn) 2025-04-15 21:01:15 [Previous line repeated 1 more time] 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 825, in _apply 2025-04-15 21:01:15 param_applied = fn(param) 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1150, in convert 2025-04-15 21:01:15 return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None, non_blocking) 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/cuda/__init__.py", line 302, in _lazy_init 2025-04-15 21:01:15 torch._C._cuda_init() 2025-04-15 21:01:15 RuntimeError: Found no NVIDIA driver on your system. Please check that you have an NVIDIA GPU and installed a driver from http://www.nvidia.com/Download/index.aspx 2025-04-15 21:01:18 2025-04-15 21:01:18,098 - cv2box - INFO - Use default multi mode: multi-thread, or you can set env 'CV_MULTI_MODE' to multi-process/torch-processC:\>nvcc --version nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2024 NVIDIA Corporation Built on Thu_Sep_12_02:55:00_Pacific_Daylight_Time_2024 Cuda compilation tools, release 12.6, V12.6.77 Build cuda_12.6.r12.6/compiler.34841621_0 C:\>nvidia-smi Tue Apr 15 20:51:35 2025 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 572.83 Driver Version: 572.83 CUDA Version: 12.8 | |-----------------------------------------+------------------------+----------------------+ | GPU Name Driver-Model | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+========================+======================| | 0 NVIDIA GeForce RTX 2080 Ti WDDM | 00000000:01:00.0 On | N/A | | 27% 35C P8 14W / 250W | 1183MiB / 22528MiB | 3% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+ +-----------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage |已确认过所有的Issues。 已使用最新的docker-compose.yml windows：windows11 24H2 Docker desktop：4.35.1 Nodejs：18.20.8 WSL： 2.4.13
按这个来，大概率是cuda 和驱动太新了。 https://www.bilibili.com/video/BV1T4dBYNEA3/
docker-compose-linux.yml下面的heygem-gen-video改成这样就行了，主要是devices那里，加上 - driver: nvidia count: all
heygem-gen-video:
image: guiji2025/heygem.ai
container_name: heygem-gen-video
restart: always
runtime: nvidia
privileged: true
volumes:
- ~/heygem_data/face2face:/code/data
environment:
- NVIDIA_VISIBLE_DEVICES=all
PYTORCH_CUDA_ALLOC_CONF=max_split_size_mb:512
deploy:
resources:
reservations:
devices:
- driver: nvidia
count: all
capabilities: [gpu]
shm_size: '8g'
ports:
- '8383:8383'
command: python /code/app_local.py
networks:
- ai_network
你好，如题。我已正确安装nvidia显卡和驱动，但依然其实找不到驱动，请问有无办法解决？谢谢。 heygem-gen-video反复重新启动，log提示如下。
2025-04-15 21:01:15 Traceback (most recent call last): 2025-04-15 21:01:15 File "/code/app_local.py", line 231, in <module> 2025-04-15 21:01:15 TransDhTask.instance() 2025-04-15 21:01:15 File "trans_dh_service.py", line 1207, in trans_dh_service.TransDhTask.instance 2025-04-15 21:01:15 File "trans_dh_service.py", line 1189, in trans_dh_service.TransDhTask.__init__ 2025-04-15 21:01:15 File "compute_ctc_att_bnf.py", line 130, in compute_ctc_att_bnf.load_ppg_model 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1152, in to 2025-04-15 21:01:15 return self._apply(convert) 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 802, in _apply 2025-04-15 21:01:15 module._apply(fn) 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 802, in _apply 2025-04-15 21:01:15 module._apply(fn) 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 802, in _apply 2025-04-15 21:01:15 module._apply(fn) 2025-04-15 21:01:15 [Previous line repeated 1 more time] 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 825, in _apply 2025-04-15 21:01:15 param_applied = fn(param) 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1150, in convert 2025-04-15 21:01:15 return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None, non_blocking) 2025-04-15 21:01:15 File "/usr/local/python3/lib/python3.8/site-packages/torch/cuda/__init__.py", line 302, in _lazy_init 2025-04-15 21:01:15 torch._C._cuda_init() 2025-04-15 21:01:15 RuntimeError: Found no NVIDIA driver on your system. Please check that you have an NVIDIA GPU and installed a driver from http://www.nvidia.com/Download/index.aspx 2025-04-15 21:01:18 2025-04-15 21:01:18,098 - cv2box - INFO - Use default multi mode: multi-thread, or you can set env 'CV_MULTI_MODE' to multi-process/torch-processC:\>nvcc --version nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2024 NVIDIA Corporation Built on Thu_Sep_12_02:55:00_Pacific_Daylight_Time_2024 Cuda compilation tools, release 12.6, V12.6.77 Build cuda_12.6.r12.6/compiler.34841621_0 C:\>nvidia-smi Tue Apr 15 20:51:35 2025 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 572.83 Driver Version: 572.83 CUDA Version: 12.8 | |-----------------------------------------+------------------------+----------------------+ | GPU Name Driver-Model | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+========================+======================| | 0 NVIDIA GeForce RTX 2080 Ti WDDM | 00000000:01:00.0 On | N/A | | 27% 35C P8 14W / 250W | 1183MiB / 22528MiB | 3% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+ +-----------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage |已确认过所有的Issues。 已使用最新的docker-compose.yml windows：windows11 24H2 Docker desktop：4.35.1 Nodejs：18.20.8 WSL： 2.4.13
按这个来，大概率是cuda 和驱动太新了。 https://www.bilibili.com/video/BV1T4dBYNEA3/
非常感谢，我先尝试一下。
docker-compose-linux.yml下面的heygem-gen-video改成这样就行了，主要是devices那里，加上 - driver: nvidia count: all
heygem-gen-video: image: guiji2025/heygem.ai container_name: heygem-gen-video restart: always runtime: nvidia privileged: true volumes: - ~/heygem_data/face2face:/code/data environment: - NVIDIA_VISIBLE_DEVICES=all PYTORCH_CUDA_ALLOC_CONF=max_split_size_mb:512 deploy: resources: reservations: devices: - driver: nvidia count: all capabilities: [gpu] shm_size: '8g' ports: - '8383:8383' command: python /code/app_local.py networks: - ai_network
非常感谢，我先尝试一下。
docker-compose-linux.yml下面的heygem-gen-video改成这样就行了，主要是devices那里，加上 - driver: nvidia count: all
heygem-gen-video: image: guiji2025/heygem.ai container_name: heygem-gen-video restart: always runtime: nvidia privileged: true volumes: - ~/heygem_data/face2face:/code/data environment: - NVIDIA_VISIBLE_DEVICES=all PYTORCH_CUDA_ALLOC_CONF=max_split_size_mb:512 deploy: resources: reservations: devices: - driver: nvidia count: all capabilities: [gpu] shm_size: '8g' ports: - '8383:8383' command: python /code/app_local.py networks: - ai_network非常感谢，我先尝试一下。
我在朋友机器了 2080 cuda 12.8也测试通过了。你可以先按这个试试，现在CUDA也不是问题
docker-compose-linux.yml下面的heygem-gen-video改成这样就行了，主要是devices那里，加上 - driver: nvidia count: all
heygem-gen-video: image: guiji2025/heygem.ai container_name: heygem-gen-video restart: always runtime: nvidia privileged: true volumes: - ~/heygem_data/face2face:/code/data environment: - NVIDIA_VISIBLE_DEVICES=all PYTORCH_CUDA_ALLOC_CONF=max_split_size_mb:512 deploy: resources: reservations: devices: - driver: nvidia count: all capabilities: [gpu] shm_size: '8g' ports: - '8383:8383' command: python /code/app_local.py networks: - ai_network非常感谢，我先尝试一下。
我在朋友机器了 2080 cuda 12.8也测试通过了。你可以先按这个试试，现在CUDA也不是问题
非常感谢。有事迟了点，但是你的方法确实可行。我修改了windows的配置，也一样可行。太谢谢了。
docker-compose-linux.yml下面的heygem-gen-video改成这样就行了，主要是devices那里，加上 - driver: nvidia count: all
heygem-gen-video: image: guiji2025/heygem.ai container_name: heygem-gen-video restart: always runtime: nvidia privileged: true volumes: - ~/heygem_data/face2face:/code/data environment: - NVIDIA_VISIBLE_DEVICES=all PYTORCH_CUDA_ALLOC_CONF=max_split_size_mb:512 deploy: resources: reservations: devices: - driver: nvidia count: all capabilities: [gpu] shm_size: '8g' ports: - '8383:8383' command: python /code/app_local.py networks: - ai_network
win10系统，虽然不理解为什么要加这个，但是确实可行
harbor 上 官方放的这个 heygem.ai 的镜像不再包含cuda；你们如果单独跑就用 推荐用这个历史版本：guiji2025/heygem.ai:0.0.7_sdk_slim
要么就带上那个tts的放一个pod里
