使用 官方代码微调(SFT推荐) ，根据sh文件微调时无法使用gpu

[kirakiradokidok]

运行过程中日志报告如下：
[INFO] [launch.py:139:main] 0 NV_LIBNCCL_DEV_PACKAGE=libnccl-dev=2.21.5-1+cuda12.4
[INFO] [launch.py:139:main] 0 NV_LIBNCCL_DEV_PACKAGE_VERSION=2.21.5-1
[INFO] [launch.py:139:main] 0 NCCL_VERSION=2.21.5-1
[INFO] [launch.py:139:main] 0 NV_LIBNCCL_DEV_PACKAGE_NAME=libnccl-dev
[INFO] [launch.py:139:main] 0 NV_LIBNCCL_PACKAGE=libnccl2=2.21.5-1+cuda12.4
[INFO] [launch.py:139:main] 0 NV_LIBNCCL_PACKAGE_NAME=libnccl2
[INFO] [launch.py:139:main] 0 NV_LIBNCCL_PACKAGE_VERSION=2.21.5-1
[WARNING] [real_accelerator.py:174:get_accelerator] Setting accelerator to CPU. If you have GPU or other accelerator, we were unable to detect it.
这是我的设备环境：
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2024 NVIDIA Corporation
Built on Thu_Mar_28_02:18:24_PDT_2024
Cuda compilation tools, release 12.4, V12.4.131
Build cuda_12.4.r12.4/compiler.34097967_0
Thu Dec 12 11:02:20 2024
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.120 Driver Version: 550.120 CUDA Version: 12.4 |
|-----------------------------------------+------------------------+----------------------+
| GPU Name Persistence-M | 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 4090 D On | 00000000:28:00.0 Off | Off |
| 0% 29C P8 21W / 425W | 4MiB / 24564MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=========================================================================================|
| No running processes found |
+-----------------------------------------------------------------------------------------+