Problem with training recognition model for English text #14753

Ibrah-N · 2025-02-24T05:17:06Z

Ibrah-N
Feb 24, 2025

I copied the below code from the paddleOCR website but don't working by saying no module named paddle.

!git clone https://github.com/PaddlePaddle/PaddleOCR.git

%cd PaddleOCR/

Download the pre-trained model of en_PP-OCRv3

!wget -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_rec_train.tar

Decompress model parameters

%cd pretrain_models
!tar -xf en_PP-OCRv3_rec_train.tar && rm -rf en_PP-OCRv3_rec_train.tar

%cd /content/PaddleOCR

#specify the single card training(Long training time, not recommended)
!python3 tools/train.py -c configs/rec/PP-OCRv3/en_PP-OCRv3_rec.yml -o Global.pretrained_model=en_PP-OCRv3_rec_train/best_accuracy

##################
Error
##################
Traceback (most recent call last):
File "/content/PaddleOCR/tools/train.py", line 31, in
from ppocr.data import build_dataloader, set_signal_handlers
File "/content/PaddleOCR/ppocr/data/init.py", line 24, in
import paddle
ModuleNotFoundError: No module named 'paddle'

GreatV · 2025-02-24T05:22:03Z

GreatV
Feb 24, 2025
Maintainer

The error you're encountering, ModuleNotFoundError: No module named 'paddle', indicates that the paddle library, which is a dependency for PaddleOCR, is not installed in your environment. To resolve this issue, you need to install the PaddlePaddle framework before running the PaddleOCR training script.

Here’s how you can fix the issue:

Install PaddlePaddle:
- You need to install the PaddlePaddle framework, which is the core dependency for PaddleOCR. You can install it using pip. Depending on your environment (CPU or GPU), you can choose the appropriate installation command.
For CPU:
```
!python -m pip install paddlepaddle==3.0.0rc1 -i https://www.paddlepaddle.org.cn/packages/stable/cpu/
```
For GPU (with CUDA support):
```
! python -m pip install paddlepaddle-gpu==3.0.0rc1 -i https://www.paddlepaddle.org.cn/packages/stable/cu118/
```
Verify Installation:
- After installing PaddlePaddle, you can verify the installation by running the following Python code:
```
import paddle
paddle.utils.run_check()
```
If the installation is successful, you should see a message indicating that PaddlePaddle is installed correctly.

Run the Training Script Again:

Once PaddlePaddle is installed, you can run the training script again:

!python3 tools/train.py -c configs/rec/PP-OCRv3/en_PP-OCRv3_rec.yml -o Global.pretrained_model=en_PP-OCRv3_rec_train/best_accuracy

Additional Notes:

Ensure that you are using a compatible version of Python (3.6 or later) and that your environment has all the necessary dependencies installed.
If you are using a GPU, make sure that the correct version of CUDA and cuDNN is installed and properly configured.

By following these steps, you should be able to resolve the ModuleNotFoundError and proceed with training your English text recognition model using PaddleOCR.

Response generated by 🤖 feifei-bot | deepseek-chat

⚠️ This is an AI-generated response. Please verify any critical information.

6 replies

GreatV Feb 24, 2025
Maintainer

please refer to https://www.paddlepaddle.org.cn/install/quick

Ibrah-N Feb 24, 2025
Author

The Script run successfully but don't give any training results just showing the given text.

code:
!python3 tools/train.py -c configs/rec/PP-OCRv3/en_PP-OCRv3_rec.yml -o Global.pretrained_model=pretrain_models/en_PP-OCRv3_rec_train/best_accuracy.pdparams

Error:
[2025/02/24 12:46:05] ppocr WARNING: Skipping import of the encryption module.
[2025/02/24 12:46:06] ppocr INFO: Architecture :
[2025/02/24 12:46:06] ppocr INFO: Backbone :
[2025/02/24 12:46:06] ppocr INFO: last_conv_stride : [1, 2]
[2025/02/24 12:46:06] ppocr INFO: last_pool_kernel_size : [2, 2]
[2025/02/24 12:46:06] ppocr INFO: last_pool_type : avg
[2025/02/24 12:46:06] ppocr INFO: name : MobileNetV1Enhance
[2025/02/24 12:46:06] ppocr INFO: scale : 0.5
[2025/02/24 12:46:06] ppocr INFO: Head :
[2025/02/24 12:46:06] ppocr INFO: head_list :
[2025/02/24 12:46:06] ppocr INFO: CTCHead :
[2025/02/24 12:46:06] ppocr INFO: Head :
[2025/02/24 12:46:06] ppocr INFO: fc_decay : 1e-05
[2025/02/24 12:46:06] ppocr INFO: Neck :
[2025/02/24 12:46:06] ppocr INFO: depth : 2
[2025/02/24 12:46:06] ppocr INFO: dims : 64
[2025/02/24 12:46:06] ppocr INFO: hidden_dims : 120
[2025/02/24 12:46:06] ppocr INFO: name : svtr
[2025/02/24 12:46:06] ppocr INFO: use_guide : True
[2025/02/24 12:46:06] ppocr INFO: SARHead :
[2025/02/24 12:46:06] ppocr INFO: enc_dim : 512
[2025/02/24 12:46:06] ppocr INFO: max_text_length : 25
[2025/02/24 12:46:06] ppocr INFO: name : MultiHead
[2025/02/24 12:46:06] ppocr INFO: Transform : None
[2025/02/24 12:46:06] ppocr INFO: algorithm : SVTR_LCNet
[2025/02/24 12:46:06] ppocr INFO: model_type : rec
[2025/02/24 12:46:06] ppocr INFO: Eval :
[2025/02/24 12:46:06] ppocr INFO: dataset :
[2025/02/24 12:46:06] ppocr INFO: data_dir : ./train_data
[2025/02/24 12:46:06] ppocr INFO: label_file_list : ['./train_data/rec/val_list.txt']
[2025/02/24 12:46:06] ppocr INFO: name : SimpleDataSet
[2025/02/24 12:46:06] ppocr INFO: transforms :
[2025/02/24 12:46:06] ppocr INFO: DecodeImage :
[2025/02/24 12:46:06] ppocr INFO: channel_first : False
[2025/02/24 12:46:06] ppocr INFO: img_mode : BGR
[2025/02/24 12:46:06] ppocr INFO: MultiLabelEncode : None
[2025/02/24 12:46:06] ppocr INFO: RecResizeImg :
[2025/02/24 12:46:06] ppocr INFO: image_shape : [3, 48, 320]
[2025/02/24 12:46:06] ppocr INFO: KeepKeys :
[2025/02/24 12:46:06] ppocr INFO: keep_keys : ['image', 'label_ctc', 'label_sar', 'length', 'valid_ratio']
[2025/02/24 12:46:06] ppocr INFO: loader :
[2025/02/24 12:46:06] ppocr INFO: batch_size_per_card : 2
[2025/02/24 12:46:06] ppocr INFO: drop_last : False
[2025/02/24 12:46:06] ppocr INFO: num_workers : 4
[2025/02/24 12:46:06] ppocr INFO: shuffle : False
[2025/02/24 12:46:06] ppocr INFO: Global :
[2025/02/24 12:46:06] ppocr INFO: cal_metric_during_train : True
[2025/02/24 12:46:06] ppocr INFO: character_dict_path : ppocr/utils/en_dict.txt
[2025/02/24 12:46:06] ppocr INFO: checkpoints : None
[2025/02/24 12:46:06] ppocr INFO: debug : False
[2025/02/24 12:46:06] ppocr INFO: distributed : False
[2025/02/24 12:46:06] ppocr INFO: epoch_num : 500
[2025/02/24 12:46:06] ppocr INFO: eval_batch_step : [0, 2000]
[2025/02/24 12:46:06] ppocr INFO: infer_img : doc/imgs_words/ch/word_1.jpg
[2025/02/24 12:46:06] ppocr INFO: infer_mode : False
[2025/02/24 12:46:06] ppocr INFO: log_smooth_window : 20
[2025/02/24 12:46:06] ppocr INFO: max_text_length : 25
[2025/02/24 12:46:06] ppocr INFO: pretrained_model : pretrain_models/en_PP-OCRv3_rec_train/best_accuracy.pdparams
[2025/02/24 12:46:06] ppocr INFO: print_batch_step : 2
[2025/02/24 12:46:06] ppocr INFO: save_epoch_step : 3
[2025/02/24 12:46:06] ppocr INFO: save_inference_dir : None
[2025/02/24 12:46:06] ppocr INFO: save_model_dir : ./output/v3_en_mobile
[2025/02/24 12:46:06] ppocr INFO: save_res_path : ./output/rec/predicts_ppocrv3_en.txt
[2025/02/24 12:46:06] ppocr INFO: use_gpu : True
[2025/02/24 12:46:06] ppocr INFO: use_space_char : True
[2025/02/24 12:46:06] ppocr INFO: use_visualdl : False
[2025/02/24 12:46:06] ppocr INFO: Loss :
[2025/02/24 12:46:06] ppocr INFO: loss_config_list :
[2025/02/24 12:46:06] ppocr INFO: CTCLoss : None
[2025/02/24 12:46:06] ppocr INFO: SARLoss : None
[2025/02/24 12:46:06] ppocr INFO: name : MultiLoss
[2025/02/24 12:46:06] ppocr INFO: Metric :
[2025/02/24 12:46:06] ppocr INFO: ignore_space : False
[2025/02/24 12:46:06] ppocr INFO: main_indicator : acc
[2025/02/24 12:46:06] ppocr INFO: name : RecMetric
[2025/02/24 12:46:06] ppocr INFO: Optimizer :
[2025/02/24 12:46:06] ppocr INFO: beta1 : 0.9
[2025/02/24 12:46:06] ppocr INFO: beta2 : 0.999
[2025/02/24 12:46:06] ppocr INFO: lr :
[2025/02/24 12:46:06] ppocr INFO: learning_rate : 0.001
[2025/02/24 12:46:06] ppocr INFO: name : Cosine
[2025/02/24 12:46:06] ppocr INFO: warmup_epoch : 5
[2025/02/24 12:46:06] ppocr INFO: name : Adam
[2025/02/24 12:46:06] ppocr INFO: regularizer :
[2025/02/24 12:46:06] ppocr INFO: factor : 3e-05
[2025/02/24 12:46:06] ppocr INFO: name : L2
[2025/02/24 12:46:06] ppocr INFO: PostProcess :
[2025/02/24 12:46:06] ppocr INFO: name : CTCLabelDecode
[2025/02/24 12:46:06] ppocr INFO: Train :
[2025/02/24 12:46:06] ppocr INFO: dataset :
[2025/02/24 12:46:06] ppocr INFO: data_dir : ./train_data/
[2025/02/24 12:46:06] ppocr INFO: ext_op_transform_idx : 1
[2025/02/24 12:46:06] ppocr INFO: label_file_list : ['./train_data/rec/rec_gt_train.txt']
[2025/02/24 12:46:06] ppocr INFO: name : SimpleDataSet
[2025/02/24 12:46:06] ppocr INFO: transforms :
[2025/02/24 12:46:06] ppocr INFO: DecodeImage :
[2025/02/24 12:46:06] ppocr INFO: channel_first : False
[2025/02/24 12:46:06] ppocr INFO: img_mode : BGR
[2025/02/24 12:46:06] ppocr INFO: RecConAug :
[2025/02/24 12:46:06] ppocr INFO: ext_data_num : 2
[2025/02/24 12:46:06] ppocr INFO: image_shape : [48, 320, 3]
[2025/02/24 12:46:06] ppocr INFO: max_text_length : 25
[2025/02/24 12:46:06] ppocr INFO: prob : 0.5
[2025/02/24 12:46:06] ppocr INFO: RecAug : None
[2025/02/24 12:46:06] ppocr INFO: MultiLabelEncode : None
[2025/02/24 12:46:06] ppocr INFO: RecResizeImg :
[2025/02/24 12:46:06] ppocr INFO: image_shape : [3, 48, 320]
[2025/02/24 12:46:06] ppocr INFO: KeepKeys :
[2025/02/24 12:46:06] ppocr INFO: keep_keys : ['image', 'label_ctc', 'label_sar', 'length', 'valid_ratio']
[2025/02/24 12:46:06] ppocr INFO: loader :
[2025/02/24 12:46:06] ppocr INFO: batch_size_per_card : 2
[2025/02/24 12:46:06] ppocr INFO: drop_last : True
[2025/02/24 12:46:06] ppocr INFO: num_workers : 4
[2025/02/24 12:46:06] ppocr INFO: shuffle : True
[2025/02/24 12:46:06] ppocr INFO: profiler_options : None
[2025/02/24 12:46:06] ppocr INFO: train with paddle 2.6.2 and device Place(gpu:0)
[2025/02/24 12:46:06] ppocr INFO: Initialize indexs of datasets:['./train_data/rec/rec_gt_train.txt']
[2025/02/24 12:46:06] ppocr INFO: Initialize indexs of datasets:['./train_data/rec/val_list.txt']
W0224 12:46:06.151784 6264 gpu_resources.cc:119] Please NOTE: device: 0, GPU Compute Capability: 7.5, Driver API Version: 12.4, Runtime API Version: 11.8
W0224 12:46:06.152729 6264 gpu_resources.cc:164] device: 0, cuDNN Version: 9.2.
[2025/02/24 12:46:06] ppocr INFO: train dataloader has 3 iters
[2025/02/24 12:46:06] ppocr INFO: valid dataloader has 3 iters
[2025/02/24 12:46:06] ppocr INFO: load pretrain successful from pretrain_models/en_PP-OCRv3_rec_train/best_accuracy
[2025/02/24 12:46:06] ppocr INFO: During the training process, after the 0th iteration, an evaluation is run every 2000 iterations
_---------> Data {'img_path': './train_data/rec/train/crop_0_4a46e8f4-images-0.jpg', 'label': 'OM HP FLARE HEADER'}
_---------> Data {'img_path': './train_data/rec/train/crop_2_4a46e8f4-images-0.jpg', 'label': '14-01-08-0637-1'}
_---------> Data {'img_path': './train_data/rec/train/crop_4_4a46e8f4-images-0.jpg', 'label': '42"-VG-9734-011503-X'}
_---------> Data {'img_path': './train_data/rec/train/crop_1_4a46e8f4-images-0.jpg', 'label': '14-01-08-0637-1'}
_---------> Data {'img_path': './train_data/rec/train/crop_5_4a46e8f4-images-0.jpg', 'label': '3/4"-VG-9732-011503-X'}
_---------> Data {'img_path': './train_data/rec/train/crop_3_4a46e8f4-images-0.jpg', 'label': 'O SLUG CATCHER'}

C++ Traceback (most recent call last):

No stack trace in paddle, may be caused by external reasons.

Error Message Summary:

FatalError: Segmentation fault is detected by the operating system.
[TimeInfo: *** Aborted at 1740401168 (unix time) try "date -d @1740401168" if you are using GNU date ***]
[SignalInfo: *** SIGSEGV (@0x0) received by PID 6264 (TID 0x7f28be201000) from PID 0 ***]

GreatV Feb 24, 2025
Maintainer

Unable to determine the error, please refer to https://paddlepaddle.github.io/PaddleOCR/main/en/ppocr/model_train/recognition.html to complete the training.

Ibrah-N Feb 24, 2025
Author

Sir I followed the exact steps mentioned in this link. But still getting this error.

GreatV Feb 24, 2025
Maintainer

Is there a more detailed and complete error message?

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Problem with training recognition model for English text #14753

Uh oh!

{{title}}

Uh oh!

Replies: 1 comment 6 replies

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{editor}}'s edit

{{editor}}'s edit

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{editor}}'s edit

{{editor}}'s edit

Uh oh!

Uh oh!

{{title}}

Uh oh!

Select a reply

Uh oh!

Problem with training recognition model for English text #14753

Uh oh!

Ibrah-N Feb 24, 2025

Download the pre-trained model of en_PP-OCRv3

Decompress model parameters

Replies: 1 comment · 6 replies

Uh oh!

Uh oh!

GreatV Feb 24, 2025 Maintainer

Additional Notes:

Uh oh!

GreatV Feb 24, 2025 Maintainer

Uh oh!

Ibrah-N Feb 24, 2025 Author

C++ Traceback (most recent call last):

Error Message Summary:

Uh oh!

GreatV Feb 24, 2025 Maintainer

Uh oh!

Uh oh!

Ibrah-N Feb 24, 2025 Author

Uh oh!

GreatV Feb 24, 2025 Maintainer

Ibrah-N
Feb 24, 2025

Replies: 1 comment 6 replies

GreatV
Feb 24, 2025
Maintainer

GreatV Feb 24, 2025
Maintainer

Ibrah-N Feb 24, 2025
Author

GreatV Feb 24, 2025
Maintainer

Ibrah-N Feb 24, 2025
Author

GreatV Feb 24, 2025
Maintainer