成功进行Mask-RT-DETR-S模型训练后，模型评估出现报错

[SXleader]

Checklist:

• 查找历史相关issue寻求解答
• 翻阅FAQ
• 翻阅PaddleX 文档
• 确认bug是否在新版本里还未修复

描述问题

复现

1. 您是否已经正常运行我们提供的教程？
   是
2. 您是否在教程的基础上修改代码内容？还请您提供运行的代码
   否
3. 您使用的数据集是？
   自有数据集，coco格式
4. 请提供您出现的报错信息及相关log
   (paddle) PS E:\study\PaddleX> python main.py -c E:/study/PaddleX/paddlex/configs/modules/instance_segmentation/Mask-RT-DETR-S.yaml -o Global.mode=evaluate -o Global.dataset_dir=F:/study_datasets/Point1_dataset/datasets_coco/dataset_dir
   ['D:\anaconda3\envs\paddle\python.exe', 'tools/eval.py', '--config', 'C:\Users\admin\.paddlex\tmp4n7qt6z7\instancesegmodel_Mask-RT-DETR-S.yml']
   ��Ϣ: ���ṩ��ģʽ�޷��ҵ��ļ���
   [03/06 21:02:32] ppdet.utils.checkpoint INFO: Skipping import of the encryption module.
   Error when construct ImageFolder instance from yaml config
   Traceback (most recent call last):
   File "tools/eval.py", line 215, in
   main()
   File "tools/eval.py", line 164, in main
   cfg = load_config(FLAGS.config)
   File "E:\study\PaddleX\paddlex\repo_manager\repos\PaddleDetection\ppdet\core\workspace.py", line 124, in load_config
   cfg = _load_config_with_base(file_path)
   File "E:\study\PaddleX\paddlex\repo_manager\repos\PaddleDetection\ppdet\core\workspace.py", line 88, in load_config_with_base
   file_cfg = yaml.load(f, Loader=yaml.Loader)
   File "D:\anaconda3\envs\paddle\lib\site-packages\yaml_init.py", line 81, in load
   return loader.get_single_data()
   File "D:\anaconda3\envs\paddle\lib\site-packages\yaml\constructor.py", line 51, in get_single_data
   return self.construct_document(node)
   File "D:\anaconda3\envs\paddle\lib\site-packages\yaml\constructor.py", line 60, in construct_document
   for dummy in generator:
   File "D:\anaconda3\envs\paddle\lib\site-packages\yaml\constructor.py", line 413, in construct_yaml_map
   value = self.construct_mapping(node)
   File "D:\anaconda3\envs\paddle\lib\site-packages\yaml\constructor.py", line 218, in construct_mapping
   return super().construct_mapping(node, deep=deep)
   File "D:\anaconda3\envs\paddle\lib\site-packages\yaml\constructor.py", line 143, in construct_mapping
   value = self.construct_object(value_node, deep=deep)
   File "D:\anaconda3\envs\paddle\lib\site-packages\yaml\constructor.py", line 100, in construct_object
   data = constructor(self, node)
   File "E:\study\PaddleX\paddlex\repo_manager\repos\PaddleDetection\ppdet\core\config\yaml_helpers.py", line 45, in python_constructor
   raise ex
   File "E:\study\PaddleX\paddlex\repo_manager\repos\PaddleDetection\ppdet\core\config\yaml_helpers.py", line 41, in python_constructor
   return cls(**kwargs)
   TypeError: new() got multiple values for argument 'cls'
   Traceback (most recent call last):
   File "E:\study\PaddleX\paddlex\utils\result_saver.py", line 29, in wrap
   result = func(self, *args, **kwargs)
   File "E:\study\PaddleX\paddlex\engine.py", line 44, in run
   return self._model.evaluate()
   File "E:\study\PaddleX\paddlex\model.py", line 105, in evaluate
   return evaluator.evaluate()
   File "E:\study\PaddleX\paddlex\modules\base\evaluator.py", line 118, in evaluate
   evaluate_result = self.pdx_model.evaluate(**self.get_eval_kwargs())
   File "E:\study\PaddleX\paddlex\repo_apis\PaddleDetection_api\instance_seg\model.py", line 190, in evaluate
   cp = self.runner.evaluate(config_path, cli_args, device, ips)
   File "E:\study\PaddleX\paddlex\repo_apis\PaddleDetection_api\instance_seg\runner.py", line 82, in evaluate
   cp = self.run_cmd(
   File "E:\study\PaddleX\paddlex\repo_apis\base\runner.py", line 356, in run_cmd
   raise CalledProcessError(
   paddlex.utils.errors.others.CalledProcessError: Command ['D:\anaconda3\envs\paddle\python.exe', 'tools/eval.py', '--config', 'C:\Users\admin\.paddlex\tmp4n7qt6z7\instancesegmodel_Mask-RT-DETR-S.yml'] returned non-zero exit status 1.

环境

1. 请提供您使用的PaddlePaddle和PaddleX的版本号
   3.0.0b2 3.0.0rc0
2. 请提供您使用的操作系统信息，如Linux/Windows/MacOS
   Windows
3. 请问您使用的Python版本是？
   3.8
4. 请问您使用的CUDA/cuDNN的版本号是？
   11.8 8.9.2.26

[SXleader]

把paddlepaddle版本修改为3.0.0rc1还是有报错

[Bobholamovic]

请问有修改过配置文件吗

[SXleader]

请问有修改过配置文件吗

您好，修改配置文件也只是修改一下路径，没有在其他地方进行修改
Global:
model: Mask-RT-DETR-S
mode: check_dataset # check_dataset/train/evaluate/predict
dataset_dir: "F:/study_datasets/Point1_dataset/datasets_coco/dataset_dir"
device: gpu:0
output: "output/Mask-RT-DETR-S"

CheckDataset:
convert:
enable: False
src_dataset_type: null
split:
enable: False
train_percent: null
val_percent: null

Train:
num_classes: 1
epochs_iters: 40
batch_size: 1
learning_rate: 0.00005
pretrain_weight_path: E:/study/PaddleX/Mask-RT-DETR-S_pretrained.pdparams
warmup_steps: 100
resume_path: null
log_interval: 10
eval_interval: 1

Evaluate:
weight_path: "output/Mask-RT-DETR-S/best_model/best_model.pdparams"
log_interval: 10

Export:
weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/Mask-RT-DETR-S_pretrained.pdparams

Predict:
batch_size: 1
model_dir: "output/best_model/inference"
input: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_instance_segmentation_004.png"
kernel_option:
run_mode: paddle
这是我现在的配置文件

[BluebirdStory]

看起来是ppdet相关的config出了问题。传入的config在python new的时候指定了多个类。
这个config的位置在：paddlex/repo_apis/PaddleDetection_api/configs/Mask-RT-DETR-S.yaml。但是我检查了官方的config，并没有问题。
需要你配合debug，你在这个位置：

打个断点，看看kwargs到低是什么？然后我们再看下一步计划。

如果是子进程调用，断点可能无效，如果是这种情况的下，你直接把kwargs打印出来。