Docker 镜像无法正常使用 xformers

镜像：registry.cn-hangzhou.aliyuncs.com/go-to-mirror/akegarasu_lora-scripts:latest

报错代码如下：

```shell
trainer | 10:59:00-563925 INFO Training started with config file /
trainer | 训练开始，使用配置文件:
trainer | /app/lora-scripts/config/autosave/20250406-105900.toml
trainer | 10:59:00-565217 INFO Using GPU(s) / 使用 GPU: ['0', '1']
trainer | 10:59:00-565963 INFO Task a06e759e-32a5-4b38-8cdb-1c5db6118556 created
trainer | [WARNING | xformers ]: WARNING[XFORMERS]: xFormers can't load C++/CUDA extensions. xFormers was built for:
trainer | PyTorch 2.4.1+cu118 with CUDA 1108 (you have 2.4.0a0+3bcc3cddb5.nv24.07)
trainer | Python 3.10.15 (you have 3.10.12)
trainer | Please reinstall xformers (see https://github.com/facebookresearch/xformers#installing-xformers)
trainer | Memory-efficient attention, SwiGLU, sparse and more won't be available.
trainer | Set XFORMERS_MORE_DETAILS=1 for more details
trainer | [WARNING | xformers ]: WARNING[XFORMERS]: xFormers can't load C++/CUDA extensions. xFormers was built for:
trainer | PyTorch 2.4.1+cu118 with CUDA 1108 (you have 2.4.0a0+3bcc3cddb5.nv24.07)
trainer | Python 3.10.15 (you have 3.10.12)
trainer | Please reinstall xformers (see https://github.com/facebookresearch/xformers#installing-xformers)
trainer | Memory-efficient attention, SwiGLU, sparse and more won't be available.
trainer | Set XFORMERS_MORE_DETAILS=1 for more details
trainer | 2025-04-06 10:59:08 INFO prepare tokenizer train_util.py:4228
trainer | 2025-04-06 10:59:08 INFO prepare tokenizer train_util.py:4228
trainer | Traceback (most recent call last):
trainer | File "/app/lora-scripts/./scripts/stable/train_network.py", line 1115, in
trainer | Traceback (most recent call last):
trainer | File "/app/lora-scripts/./scripts/stable/train_network.py", line 1115, in
trainer | trainer.train(args)trainer.train(args)
trainer |
trainer | File "/app/lora-scripts/./scripts/stable/train_network.py", line 154, in train
trainer | File "/app/lora-scripts/./scripts/stable/train_network.py", line 154, in train
trainer | tokenizer = self.load_tokenizer(args)
trainer | File "/app/lora-scripts/./scripts/stable/train_network.py", line 105, in load_tokenizer
trainer | tokenizer = self.load_tokenizer(args)
trainer | File "/app/lora-scripts/./scripts/stable/train_network.py", line 105, in load_tokenizer
trainer | tokenizer = train_util.load_tokenizer(args)
trainer | tokenizer = train_util.load_tokenizer(args) File "/app/lora-scripts/scripts/stable/library/train_util.py", line 4242, in load_tokenizer
trainer |
trainer | File "/app/lora-scripts/scripts/stable/library/train_util.py", line 4242, in load_tokenizer
trainer | tokenizer = CLIPTokenizer.from_pretrained(original_path)
trainer | File "/usr/local/lib/python3.10/dist-packages/transformers/tokenization_utils_base.py", line 2255, in from_pretrained
trainer | tokenizer = CLIPTokenizer.from_pretrained(original_path)
trainer | File "/usr/local/lib/python3.10/dist-packages/transformers/tokenization_utils_base.py", line 2255, in from_pretrained
trainer | raise EnvironmentError(
trainer | OSError: Can't load tokenizer for 'openai/clip-vit-large-patch14'. If you were trying to load it from 'https://huggingface.co/models', make sure you don't have a local directory with the same name. Otherwise, make sure 'op
enai/clip-vit-large-patch14' is the correct path to a directory containing all relevant files for a CLIPTokenizer tokenizer.
trainer | raise EnvironmentError(
trainer | OSError: Can't load tokenizer for 'openai/clip-vit-large-patch14'. If you were trying to load it from 'https://huggingface.co/models', make sure you don't have a local directory with the same name. Otherwise, make sure 'op
enai/clip-vit-large-patch14' is the correct path to a directory containing all relevant files for a CLIPTokenizer tokenizer.
trainer | E0406 11:00:19.590000 139944006431360 torch/distributed/elastic/multiprocessing/api.py:832] failed (exitcode: 1) local_rank: 0 (pid: 950) of binary: /usr/bin/python
trainer | Traceback (most recent call last):
trainer | File "/usr/lib/python3.10/runpy.py", line 196, in _run_module_as_main
trainer | return _run_code(code, main_globals, None,
trainer | File "/usr/lib/python3.10/runpy.py", line 86, in _run_code
trainer | exec(code, run_globals)
trainer | File "/usr/local/lib/python3.10/dist-packages/accelerate/commands/launch.py", line 1116, in
trainer | main()
trainer | File "/usr/local/lib/python3.10/dist-packages/accelerate/commands/launch.py", line 1112, in main
trainer | launch_command(args)
trainer | File "/usr/local/lib/python3.10/dist-packages/accelerate/commands/launch.py", line 1097, in launch_command
trainer | multi_gpu_launcher(args)
trainer | File "/usr/local/lib/python3.10/dist-packages/accelerate/commands/launch.py", line 734, in multi_gpu_launcher
trainer | distrib_run.run(args)
trainer | File "/usr/local/lib/python3.10/dist-packages/torch/distributed/run.py", line 891, in run
trainer | elastic_launch(
trainer | File "/usr/local/lib/python3.10/dist-packages/torch/distributed/launcher/api.py", line 132, in __call__
trainer | return launch_agent(self._config, self._entrypoint, list(args))
trainer | File "/usr/local/lib/python3.10/dist-packages/torch/distributed/launcher/api.py", line 263, in launch_agent
trainer | raise ChildFailedError(
trainer | torch.distributed.elastic.multiprocessing.errors.ChildFailedError:
trainer | ============================================================
trainer | ./scripts/stable/train_network.py FAILED
trainer | ------------------------------------------------------------
trainer | Failures:
trainer | [1]:
trainer | time : 2025-04-06_11:00:19
trainer | host : trainer
trainer | rank : 1 (local_rank: 1)
trainer | exitcode : 1 (pid: 951)
trainer | error_file:
trainer | traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html
trainer | ------------------------------------------------------------
trainer | Root Cause (first observed failure):
trainer | [0]:
trainer | time : 2025-04-06_11:00:19
trainer | host : trainer
trainer | rank : 0 (local_rank: 0)
trainer | exitcode : 1 (pid: 950)
trainer | error_file:
trainer | traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html
trainer | ============================================================
trainer | 11:00:20-224226 ERROR Training failed / 训练失败

```