|
使用docker进行训练仍有问题，拉取的镜像是paddlepaddle/paddle:2.6.1-gpu-cuda11.7-cudnn8.4-trt8.4
训练报错
Traceback (most recent call last):
File "/home/PaddleOCR/tools/train.py", line 208, in
main(config, device, logger, vdl_writer)
File "/home/PaddleOCR/tools/train.py", line 121, in main
model = build_model(config['Architecture'])
File "/home/PaddleOCR/ppocr/modeling/architectures/init.py", line 34, in build_model
arch = getattr(mod, name)(config)
File "/home/PaddleOCR/ppocr/modeling/architectures/distillation_model.py", line 47, in init
model = BaseModel(model_config)
File "/home/PaddleOCR/ppocr/modeling/architectures/base_model.py", line 55, in init
self.backbone = build_backbone(config["Backbone"], model_type)
File "/home/PaddleOCR/ppocr/modeling/backbones/init.py", line 74, in build_backbone
module_class = eval(module_name)(**config)
File "/home/PaddleOCR/ppocr/modeling/backbones/vqa_layoutlm.py", line 149, in init
super(LayoutXLMForSer, self).init(
File "/home/PaddleOCR/ppocr/modeling/backbones/vqa_layoutlm.py", line 65, in init
self.model = model_class(
File "/usr/local/lib/python3.10/dist-packages/paddlenlp/transformers/utils.py", line 255, in impl
init_func(self, *args, **kwargs)
TypeError: LayoutXLMForTokenClassification.init() got an unexpected keyword argument 'num_classes'
官方有提供paddlepaddle-gpu 2.5.2.post117、paddlenlp 2.5.2的docker镜像吗
|