昇腾910B服务器paddlex无法解析带有图片的pdf文件。测试环境为官网拉的镜像。 #3419
Assignees
Comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Checklist:
查找历史相关issue寻求解答
翻阅FAQ
翻阅PaddleX 文档
是
确认bug是否在新版本里还未修复
未检索到相关bug
描述问题
根据官网拉的镜像：https://paddlepaddle.github.io/PaddleX/latest/other_devices_support/paddlepaddle_install_NPU.html。使用官网教程安装了PaddleOCR、PaddleDetection、PaddleClas三个插件。在运行示例代码时无法解析带图片的pdf文件，纯文字的pdf是ok的。
复现
是
`from paddlex import create_pipeline
pipeline = create_pipeline(pipeline="layout_parsing", device="npu:1") # 将设备名修改为 npu、mlu、xpu、dcu 或 gcu
output = pipeline.predict("安全强化学习综述.pdf")
for res in output:
res.print()`
3. 您使用的数据集是？
安全强化学习综述.pdf
请提供您出现的报错信息及相关log
7eb6889ba57d /app python a.py
Using official model (RT-DETR-H_layout_17cls), the model files will be be automatically downloaded and saved in /root/.paddlex/official_models.
I0220 17:18:39.139822 4905 init.cc:236] ENV [CUSTOM_DEVICE_ROOT]=/usr/local/lib/python3.9/dist-packages/paddle_custom_device
I0220 17:18:39.139883 4905 init.cc:145] Try loading custom device libs from: [/usr/local/lib/python3.9/dist-packages/paddle_custom_device]
I0220 17:18:39.648082 4905 custom_device.cc:1099] Succeed in loading custom runtime in lib: /usr/local/lib/python3.9/dist-packages/paddle_custom_device/libpaddle-custom-npu.so
I0220 17:18:39.654825 4905 custom_kernel.cc:63] Succeed in loading 355 custom kernel(s) from loaded lib(s), will be used like native ones.
I0220 17:18:39.654999 4905 init.cc:157] Finished in LoadCustomDevice with libs_path: [/usr/local/lib/python3.9/dist-packages/paddle_custom_device]
I0220 17:18:39.655045 4905 init.cc:242] CustomDevice: npu, visible devices count: 1
Using official model (PP-OCRv4_server_det), the model files will be be automatically downloaded and saved in /root/.paddlex/official_models.
Using official model (PP-OCRv4_server_rec), the model files will be be automatically downloaded and saved in /root/.paddlex/official_models.
Using official model (SLANet_plus), the model files will be be automatically downloaded and saved in /root/.paddlex/official_models.
Using official model (LaTeX_OCR_rec), the model files will be be automatically downloaded and saved in /root/.paddlex/official_models.
Using official model (PP-OCRv4_server_seal_det), the model files will be be automatically downloaded and saved in /root/.paddlex/official_models.
Using official model (PP-OCRv4_server_rec), the model files will be be automatically downloaded and saved in /root/.paddlex/official_models.
{'layout_parsing_result': {'input_path': PosixPath('/root/.paddlex/temp/tmpexfrrpvk.png'), 'parsing_result': [{'input_path': PosixPath('/root/.paddlex/temp/tmpexfrrpvk.png'), 'layout_bbox': array([119., 81., 266., 105.], dtype=float32), 'text_without_layout': '第49卷第9期', 'layout': 'double'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpexfrrpvk.png'), 'layout_bbox': array([121., 114., 220., 133.], dtype=float32), 'text_without_layout': '2023年9月', 'layout': 'double'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpexfrrpvk.png'), 'layout_bbox': array([526., 79., 693., 103.], dtype=float32), 'text_without_layout': '自动化学报', 'layout': 'double'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpexfrrpvk.png'), 'layout_bbox': array([ 946., 81., 1065., 105.], dtype=float32), 'text_without_layout': 'Vol.49,No. 9', 'layout': 'double'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpexfrrpvk.png'), 'layout_bbox': array([480., 112., 743., 135.], dtype=float32), 'text_without_layout': 'ACTA AUTOMATICA SINICA', 'layout': 'double'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpexfrrpvk.png'), 'layout_bbox': array([ 927., 112., 1065., 137.], dtype=float32), 'text_without_layout': 'September,2023', 'layout': 'double'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpujq8fwpi.png'), 'layout_bbox': [446.26456, 195.51045, 734.45105, 230.59236], 'paragraph_title': '安全强化学习综述', 'layout': 'single'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpjlajmi3p.png'), 'layout_bbox': [446.26456, 195.51045, 734.45105, 230.59236], 'doc_title': '安全强化学习综述', 'layout': 'single'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpg7x4jv9s.png'), 'layout_bbox': [102.495705, 210.4346, 1074.9762, 565.9726], 'abstract': '安全强化学习综还\n王雪松’王荣荣’程玉虎\n摘要强化学习(Reinforcementlearning，RL)在围棋、视频游戏、导航、推荐系统等领域均取得了巨大成功.然而，许\n多强化学习算法仍然无法直接移植到真实物理环境中.这是因为在模拟场景下智能体能以不断试错的方式与环境进行交\n互，从而学习最优策略.但考虑到安全因素，很多现实世界的应用则要求限制智能体的随机探索行为.因此，安全问题成为\n强化学习从模拟到现实的一个重要挑战.近年来，许多研究致力于开发安全强化学习（Safereinforcementlearning，SRL\n算法，在确保系统性能的同时满足安全约束.本文对现有的安全强化学习算法进行全面综述，将其归为三类：修改学习过程\n修改学习目标、离线强化学习，并介绍了5大基准测试平台：SafetyGym、safe-control-gym、SafeRL-Kit、D4RL、NeoRL\n最后总结了安全强化学习在自动驾驶、机器人控制、工业过程控制、电力系统优化和医疗健康领域中的应用，并给出结论与\n展望\n关键词安全强化学习，约束马尔科夫决策过程，学习过程，学习目标，离线强化学习\n引用格式王雪松，王荣荣，程玉虎.安全强化学习综述．自动化学报，2023，49(9)：1813-1835', 'layout': 'single'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpiny8hocn.png'), 'layout_bbox': [100.64751, 298.9206, 1079.4558, 483.82715], 'text': '摘要强化学习(Reinforcementlearning,RL)在围棋、视频游戏、导航、推荐系统等领域均取得了巨大成功.然而，许\n多强化学习算法仍然无法直接移植到真实物理环境中.这是因为在模拟场景下智能体能以不断试错的方式与环境进行交\n互，从而学习最优策略.但考虑到安全因素，很多现实世界的应用则要求限制智能体的随机探索行为.因此，安全问题成为\n强化学习从模拟到现实的一个重要挑战.近年来，许多研究致力于开发安全强化学习(Safe reinforcementlearning，SRL)\n算法，在确保系统性能的同时满足安全约束.本文对现有的安全强化学习算法进行全面综述，将其归为三类：修改学习过程、\n修改学习目标、离线强化学习，并介绍了5大基准测试平台：SafetyGym、safe-control-gym、SafeRL-Kit、D4RL、NeoRL\n最后总结了安全强化学习在自动驾驶、机器人控制、工业过程控制、电力系统优化和医疗健康领域中的应用，并给出结论与\n展望', 'layout': 'single'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpexfrrpvk.png'), 'layout_bbox': array([158., 561., 335., 574.], dtype=float32), 'text_without_layout': '.1..', 'layout': 'double'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpgpt8xw5j.png'), 'layout_bbox': [378.09607, 598.7253, 802.8417, 620.6022], 'paragraph_title': 'Safe Reinforcement Learning: A Survey', 'layout': 'single'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpizunemzo.png'), 'layout_bbox': [333.0858, 643.6011, 847.9703, 662.82904], 'text': 'WANGXue-Song1\nWANG Rong-Rong CHENG Yu-Hu', 'layout': 'single'}, {'input_path': PosixPath('/root/.paddlex/temp/tmp6mclygn2.png'), 'layout_bbox': [100.53669, 684.34393, 1077.9385, 1015.9402], 'text': 'Abstract Reinforcement learning (RL)has proved a prominent success in the game of Go, video games, naviga-\ntion,recommendation systems and other fields. However, a large number of reinforcement learning algorithms can-\nnot be directly transplanted to real physical environment. This is because in the simulation scenario, the agent is\nable to interact with the environment in a trial-and-error manner to learn the optimal policy. Considering the safety\nof systems, many real-world applications require the limitation of random exploration behavior of agents. Hence.\n safety has become an essential factor for reinforcement learning from simulation to reality. In recent years, many re\n searches have been devoted to develope safe reinforcement learning (SRL) algorithms that satisfy safety constraints\nwhile ensuring system performance. This paper presents a comprehensive survey of existing SRL algorithms,which\nare divided into three categories: Modification of learning process, modification of learning objective, and offline re-\ninforcement learning.Furthermore, five experimental platforms are introduced, including Safety Gym, safe-control-\ngym, SafeRL-Kit, D4RL,and NeoRL. Lastly, the applications of SRL in the fields of autonomous driving,robot\ncontrol,industrial process control, power system optimization,and healthcare are summarized, and the conclusion\nand perspective are briefly drawn.', 'layout': 'single'}, {'input_path': PosixPath('/root/.paddlex/temp/tmptkbngj6l.png'), 'layout_bbox': [100.15541, 1024.6014, 1079.7997, 1066.0405], 'text': 'Key words Safe reinforcement learning (SRL), constrained Markov decision process (CMDP), learning process,\nlearning objective,offline reinforcement learning', 'layout': 'single'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpl0wnd1w2.png'), 'layout_bbox': [100.86576, 1074.7885, 1079.8624, 1116.3823], 'text': 'Citation Wang Xue-Song,Wang Rong-Rong,Cheng Yu-Hu. Safe reinforcement learning: A survey. Acta Automat-\nica Sinica.2023.49(9): 1813-1835', 'layout': 'single'}, {'input_path': PosixPath('/root/.paddlex/temp/tmpcal5gdmi.png'), 'layout_bbox': [101.77417, 1162.699, 568.08453, 1214.6223], 'text': '作为一种重要的机器学习方法，强化学习(Re-\ninforcementlearning，RL）采用了人类和动物学习', 'layout': 'double'}, {'input_path': PosixPath('/root/.paddlex/temp/tmp3vob7o_h.png'), 'layout_bbox': [101.77417, 1162.699, 568.08453, 1214.6223], 'paragraph_title': '作为一种重要的机器学习方法，强化学习(Re-\ninforcementlearning，RL）采用了人类和动物学习', 'layout': 'double'}, {'input_path': PosixPath('/root/.paddlex/temp/tmp4rrz8yky.png'), 'layout_bbox': [101.37287, 1250.31, 566.9036, 1481.9248], 'footnote': '收稿日期2022-08-08录用日期2023-01-11\nManuscript received August 8,2022; accepted January 11\n2023\n国家自然科学基金(62176259，61976215)，江苏省重点研发计划\n项目（BE2022095）资助\nSupported by National Natural Science Foundation of China\n(62176259, 61976215) and Key Research and Development P1\nam of Jiangsu Province (BE2022095)\n本文责任编委黎铭\nommended by Associate Editor LI Ming\n中国矿业大学信息与控制工程学院徐州221116\n of Information and Control Engineering, China Uni\nMining and Technologv.Xuzhou 221116', 'layout': 'double'}, {'input_path': PosixPath('/root/.paddlex/temp/tmp98q581b3.png'), 'layout_bbox': [611.25494, 1161.318, 1078.951, 1485.8845], 'text': '中“试错法”与“奖惩回报”的行为心理学机制，强\n调智能体在与环境的交互中学习，利用评价性的反\n馈信号实现决策的优化[．早期的强化学习主要依\n赖于人工提取特征，难以处理复杂高维状态和动作\n空间下的问题.近年来，随着计算机硬件设备性能\n的提升和神经网络学习算法的发展，深度学习由于\n其强大的表征能力和泛化性能受到了众多研究人员\n的关注[2-3].于是，将深度学习与强化学习相结合就\n成为了解决复杂环境下感知决策问题的一个可行方\n案.2016年，Google公司的研究团队DeepMind创\n新性地将具有感知能力的深度学习与具有决策能', 'layout': 'double'}], 'page_id': 1}}
Traceback (most recent call last):
File "/app/a.py", line 6, in
for res in output:
File "/home/pp_npu/PaddleX/paddlex/inference/pipelines/layout_parsing/layout_parsing.py", line 290, in predict
all_formula_res = get_formula_res(self.formula_predictor, formula_subs)
File "/home/pp_npu/PaddleX/paddlex/inference/pipelines/layout_parsing/layout_parsing.py", line 360, in get_formula_res
for res in predictor(img):
File "/home/pp_npu/PaddleX/paddlex/inference/models/base/basic_predictor.py", line 74, in call
yield from super().call(input)
File "/home/pp_npu/PaddleX/paddlex/inference/models/base/base_predictor.py", line 48, in call
for res in super().call(input):
File "/home/pp_npu/PaddleX/paddlex/inference/components/base.py", line 63, in call
for each_output in output:
File "/home/pp_npu/PaddleX/paddlex/inference/models/base/basic_predictor.py", line 78, in apply
yield from self._generate_res(self.engine(input))
File "/home/pp_npu/PaddleX/paddlex/inference/utils/process_hook.py", line 46, in wrapper
for ele in input:
File "/home/pp_npu/PaddleX/paddlex/inference/components/base.py", line 290, in call
yield from self.call(data, i + 1)
File "/home/pp_npu/PaddleX/paddlex/inference/components/base.py", line 290, in call
yield from self.call(data, i + 1)
File "/home/pp_npu/PaddleX/paddlex/inference/components/base.py", line 289, in call
for data in data_gen:
File "/home/pp_npu/PaddleX/paddlex/inference/components/base.py", line 49, in call
output = self.apply(**args)
File "/home/pp_npu/PaddleX/paddlex/inference/components/paddle_predictor/predictor.py", line 219, in apply
self.infer.apply()
File "/home/pp_npu/PaddleX/paddlex/inference/components/paddle_predictor/predictor.py", line 59, in apply
self.predictor.run()
NotImplementedError: In user code:
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/tools/export_model.py", line 37, in
main()
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/tools/export_model.py", line 33, in main
export(config)
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/ppocr/utils/export_model.py", line 376, in export
export_single_model(
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/ppocr/utils/export_model.py", line 266, in export_single_model
paddle.jit.save(model, save_path)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/decorator.py", line 232, in fun
return caller(func, *(extras + args), **kw)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/base/wrapped_decorator.py", line 40, in impl
return wrapped_func(*args, **kwargs)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/api.py", line 895, in wrapper
func(layer, path, input_spec, **configs)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/decorator.py", line 232, in fun
return caller(func, *(extras + args), **kw)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/base/wrapped_decorator.py", line 40, in impl
return wrapped_func(*args, **kwargs)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/base/dygraph/base.py", line 101, in impl
return func(*args, **kwargs)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/api.py", line 1209, in save
static_func.concrete_program_specify_input_spec(
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/program_translator.py", line 1026, in concrete_program_specify_input_spec
concrete_program, _ = self.get_concrete_program(
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/program_translator.py", line 914, in get_concrete_program
concrete_program, partial_program_layer = self._program_cache[
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/program_translator.py", line 1665, in getitem
self._caches[item_id] = self._build_once(item)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/program_translator.py", line 1611, in _build_once
concrete_program = ConcreteProgram.from_func_spec(
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/decorator.py", line 232, in fun
return caller(func, *(extras + args), **kw)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/base/wrapped_decorator.py", line 40, in impl
return wrapped_func(*args, **kwargs)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/base/dygraph/base.py", line 101, in impl
return func(*args, **kwargs)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/program_translator.py", line 1372, in from_func_spec
outputs = static_func(*inputs)
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/ppocr/modeling/architectures/base_model.py", line 98, in forward
if self.use_head:
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/convert_operators.py", line 430, in convert_ifelse
out = _run_py_ifelse(
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/convert_operators.py", line 523, in _run_py_ifelse
py_outs = true_fn() if pred else false_fn()
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/ppocr/modeling/architectures/base_model.py", line 99, in forward
x = self.head(x, targets=data)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/nn/layer/layers.py", line 1534, in call
return self._dygraph_call_func(*inputs, **kwargs)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/nn/layer/layers.py", line 1513, in _dygraph_call_func
outputs = self.forward(*inputs, **kwargs)
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/ppocr/modeling/heads/rec_latexocr_head.py", line 991, in forward
if not self.training:
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/convert_operators.py", line 430, in convert_ifelse
out = _run_py_ifelse(
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/convert_operators.py", line 523, in _run_py_ifelse
py_outs = true_fn() if pred else false_fn()
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/ppocr/modeling/heads/rec_latexocr_head.py", line 996, in forward
if self.is_export:
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/convert_operators.py", line 430, in convert_ifelse
out = _run_py_ifelse(
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/convert_operators.py", line 523, in _run_py_ifelse
py_outs = true_fn() if pred else false_fn()
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/ppocr/modeling/heads/rec_latexocr_head.py", line 997, in forward
word_pred = self.generate_export(
File "/root/.cache/paddle/to_static_tmp/43153/generate_exportvxa9srok.py", line 171, in generate_export
return _jst.Ld(_decoedby_no_grad)(_jst.Ld(self), _jst.Ld(start_tokens), _jst.Ld(seq_len), _jst.Ld(eos_token), _jst.Ld(context), _jst.Ld(temperature), _jst.Ld(filter_logits_fn), _jst.Ld(filter_thres))
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/decorator.py", line 232, in fun
return caller(func, *(extras + args), **kw)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/base/dygraph/base.py", line 397, in _decorate_function
return func(*args, **kwargs)
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/ppocr/modeling/heads/rec_latexocr_head.py", line 958, in generate_export
while i_idx < paddle.to_tensor(seq_len):
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/convert_operators.py", line 185, in convert_while_loop
_run_paddle_while(
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/convert_operators.py", line 249, in _run_paddle_while
loop_vars = while_loop(new_cond_fn, new_body_fn, loop_vars)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/static/nn/control_flow.py", line 897, in while_loop
output_vars = body(*loop_vars)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/convert_operators.py", line 231, in new_body_fn
body()
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/ppocr/modeling/heads/rec_latexocr_head.py", line 963, in generate_export
logits = self.net(x, mask=mask, context=context, seq_len=i_idx, **kwargs)[
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/nn/layer/layers.py", line 1534, in call
return self._dygraph_call_func(*inputs, **kwargs)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/nn/layer/layers.py", line 1513, in _dygraph_call_func
outputs = self.forward(*inputs, **kwargs)
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/ppocr/modeling/heads/rec_latexocr_head.py", line 785, in forward
x, intermediates = self.attn_layers(
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/nn/layer/layers.py", line 1534, in call
return self._dygraph_call_func(*inputs, **kwargs)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/nn/layer/layers.py", line 1513, in _dygraph_call_func
outputs = self.forward(*inputs, **kwargs)
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/ppocr/modeling/heads/rec_latexocr_head.py", line 623, in forward
for ind, (layer_type, (norm, block, residual_fn)) in enumerate(
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/convert_operators.py", line 189, in convert_while_loop
_run_py_while(cond, body, getter, setter)
File "/opt/conda/envs/paddlex_beta/lib/python3.10/site-packages/paddle/jit/dy2static/convert_operators.py", line 263, in _run_py_while
body()
File "/paddle/uniform_output_enabled/test_ci/PaddleX/paddlex/repo_manager/repos/PaddleOCR/ppocr/modeling/heads/rec_latexocr_head.py", line 637, in forward
if layer_type == "a":
File
环境
官网提供的昇腾镜像
宿主机欧拉系统，容器是官网提供的ubuntu
3.9
The text was updated successfully, but these errors were encountered: