运行train.py

[WGJiMin]

Traceback (most recent call last):
File "train.py", line 3, in
import pytorch_lightning as pl
File "/opt/conda/lib/python3.8/site-packages/pytorch_lightning/init.py", line 20, in
from pytorch_lightning.callbacks import Callback # noqa: E402
File "/opt/conda/lib/python3.8/site-packages/pytorch_lightning/callbacks/init.py", line 14, in
from pytorch_lightning.callbacks.base import Callback
File "/opt/conda/lib/python3.8/site-packages/pytorch_lightning/callbacks/base.py", line 26, in
from pytorch_lightning.utilities.types import STEP_OUTPUT
File "/opt/conda/lib/python3.8/site-packages/pytorch_lightning/utilities/init.py", line 18, in
from pytorch_lightning.utilities.apply_func import move_data_to_device # noqa: F401
File "/opt/conda/lib/python3.8/site-packages/pytorch_lightning/utilities/apply_func.py", line 29, in
if _compare_version("torchtext", operator.ge, "0.9.0"):
File "/opt/conda/lib/python3.8/site-packages/pytorch_lightning/utilities/imports.py", line 54, in _compare_version
pkg = importlib.import_module(package)
File "/opt/conda/lib/python3.8/importlib/init.py", line 127, in import_module
return _bootstrap._gcd_import(name[level:], package, level)
File "/opt/conda/lib/python3.8/site-packages/torchtext/init.py", line 5, in
from . import vocab
File "/opt/conda/lib/python3.8/site-packages/torchtext/vocab/init.py", line 11, in
from .vocab_factory import (
File "/opt/conda/lib/python3.8/site-packages/torchtext/vocab/vocab_factory.py", line 4, in
from torchtext._torchtext import (
ImportError: /opt/conda/lib/python3.8/site-packages/torchtext/_torchtext.so: undefined symbol: _ZNK3c104Type14isSubtypeOfExtERKSt10shared_ptrIS0_EPSo
你好，请问我运行train.py文件时出现这个错误应该如何解决？

[huai-chang]

感觉是你环境的问题。你把torchtext库卸载了试试？或者安装与当前 PyTorch 版本匹配的 torchtext？参考这个

[WGJiMin]

你好请问，我运行train.py只有输出了这些内容
/opt/conda/lib/python3.8/site-packages/horovod/common/util.py:258: UserWarning: Framework pytorch installed with version 1.10.0+cu102 but found version 1.10.0+cu113.
This can result in unexpected behavior including runtime errors.
Reinstall Horovod using pip install --no-cache-dir to build with the new version.
warnings.warn(get_version_mismatch_message(name, version, installed_version))
Global seed set to 231
[DEBUG] Instantiating data module
[DEBUG] Instantiating model
WARNING[XFORMERS]: xFormers can't load C++/CUDA extensions. xFormers was built for:
PyTorch 1.13.1+cu117 with CUDA 1107 (you have 1.10.0+cu113)
Python 3.8.16 (you have 3.8.12)
Please reinstall xformers (see https://github.com/facebookresearch/xformers#installing-xformers)
Memory-efficient attention, SwiGLU, sparse and more won't be available.
Set XFORMERS_MORE_DETAILS=1 for more details
DiffEIC: Running in eps-prediction mode
DiffusionWrapper has 865.91 M params.
making attention of type 'vanilla-xformers' with 512 in_channels
building MemoryEfficientAttnBlock with 512 in_channels...
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
making attention of type 'vanilla-xformers' with 512 in_channels
building MemoryEfficientAttnBlock with 512 in_channels...
[DEBUG] Trying to load model from: ./weight/v2-1_512-ema-pruned.ckpt
[615 keys are missing from the model (hint processing and cross connections included)]
Loading pretrained model LPIPS from /root/.cache/torch/hub/checkpoints/LPIPS_v0.1_alex-df73285e.pth
[DEBUG] Running data_module.setup('fit')
[DEBUG] Calling train_dataloader()
[DEBUG] Dataloader length: 112
GPU available: True, used: True
TPU available: False, using: 0 TPU cores
IPU available: False, using: 0 IPUs
[DEBUG] Start training...
/opt/conda/lib/python3.8/site-packages/pytorch_lightning/trainer/configuration_validator.py:280: LightningDeprecationWarning: Base LightningModule.on_train_batch_start hook signature has changed in v1.5. The dataloader_idx argument will be removed in v1.7.
rank_zero_deprecation(
/opt/conda/lib/python3.8/site-packages/pytorch_lightning/trainer/configuration_validator.py:287: LightningDeprecationWarning: Base Callback.on_train_batch_end hook signature has changed in v1.5. The dataloader_idx argument will be removed in v1.7.
rank_zero_deprecation(
/opt/conda/lib/python3.8/site-packages/pytorch_lightning/core/datamodule.py:469: LightningDeprecationWarning: DataModule.setup has already been called, so it will not be called again. In v1.6 this behavior will change to always call DataModule.setup.
rank_zero_deprecation(
LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]
gpu_partition - [InitCudaPidMap] warn:recv response msg len mismatch, rcv_len = -1, errno = 111
gpu_partition - [BackTraceStack] warn:/usr/local/lib/inais/libgpu_partition.so(+0x13b1b) [0x7f9c152c6b1b]
gpu_partition - [BackTraceStack] warn:/usr/local/lib/inais/libgpu_partition.so(cuDevicePrimaryCtxRetain+0x13b) [0x7f9c152c752f]
gpu_partition - [BackTraceStack] warn:/opt/conda/lib/python3.8/site-packages/torch/lib/libcudart-a7b20f20.so.11.0(+0x38872) [0x7f9b4baad872]
gpu_partition - [BackTraceStack] warn:/opt/conda/lib/python3.8/site-packages/torch/lib/libcudart-a7b20f20.so.11.0(+0x38a15) [0x7f9b4baada15]
gpu_partition - [BackTraceStack] warn:/opt/conda/lib/python3.8/site-packages/torch/lib/libcudart-a7b20f20.so.11.0(+0x3956c) [0x7f9b4baae56c]
Segmentation fault (core dumped)

这是我的包的版本信息：

PyTorch 和基本组件

torch==1.10.0+cu113
torchvision==0.11.1
torchaudio==0.10.0
torchtext==0.11.0
pytorch-lightning==1.5.0

DiffEIC 所需依赖

compressai==1.2.4
einops==0.8.0
kornia==0.7.0
lpips==0.1.4
numpy==1.23.1
omegaconf==2.3.0
open_clip_torch==2.22.0
openai_clip==1.0.1
opencv-contrib-python==4.8.1.78
opencv-python==4.8.1.78
opencv-python-headless==4.8.1.78
Pillow==10.4.0
pyiqa==0.1.7
thop==0.1.1.post2209072238
timm==0.9.7
tqdm==4.66.1
transformers==4.41.0
ultralytics_thop==2.0.0

xformers

xformers==0.0.14