ValueError: too many values to unpack (expected 4) #13363

aqingGitHub · 2024-07-12T07:55:09Z

aqingGitHub
Jul 12, 2024

详细信息如下，试过将Train的batch_size_per_card改成1、2、4，都无效。请教各位大神怎么解决，谢谢！
(paddle2) D:\ProgramData\PaddleOCR275>python tools/train.py -c configs/det/ch_ppocr_v2.0/ch_det_res18_db_v2.0.yml
[2024/07/12 15:49:51] ppocr INFO: Architecture :
[2024/07/12 15:49:51] ppocr INFO: Backbone :
[2024/07/12 15:49:51] ppocr INFO: disable_se : True
[2024/07/12 15:49:51] ppocr INFO: layers : 18
[2024/07/12 15:49:51] ppocr INFO: name : ResNet_vd
[2024/07/12 15:49:51] ppocr INFO: Head :
[2024/07/12 15:49:51] ppocr INFO: k : 50
[2024/07/12 15:49:51] ppocr INFO: name : DBHead
[2024/07/12 15:49:51] ppocr INFO: Neck :
[2024/07/12 15:49:51] ppocr INFO: name : DBFPN
[2024/07/12 15:49:51] ppocr INFO: out_channels : 256
[2024/07/12 15:49:51] ppocr INFO: Transform : None
[2024/07/12 15:49:51] ppocr INFO: algorithm : DB
[2024/07/12 15:49:51] ppocr INFO: model_type : det
[2024/07/12 15:49:51] ppocr INFO: Eval :
[2024/07/12 15:49:51] ppocr INFO: dataset :
[2024/07/12 15:49:51] ppocr INFO: data_dir : ./train_data
[2024/07/12 15:49:51] ppocr INFO: label_file_list : ['./train_data/det/val.txt']
[2024/07/12 15:49:51] ppocr INFO: name : SimpleDataSet
[2024/07/12 15:49:51] ppocr INFO: transforms :
[2024/07/12 15:49:51] ppocr INFO: DecodeImage :
[2024/07/12 15:49:51] ppocr INFO: channel_first : False
[2024/07/12 15:49:51] ppocr INFO: img_mode : BGR
[2024/07/12 15:49:51] ppocr INFO: DetLabelEncode : None
[2024/07/12 15:49:51] ppocr INFO: DetResizeForTest : None
[2024/07/12 15:49:51] ppocr INFO: NormalizeImage :
[2024/07/12 15:49:51] ppocr INFO: mean : [0.485, 0.456, 0.406]
[2024/07/12 15:49:51] ppocr INFO: order : hwc
[2024/07/12 15:49:51] ppocr INFO: scale : 1./255.
[2024/07/12 15:49:51] ppocr INFO: std : [0.229, 0.224, 0.225]
[2024/07/12 15:49:51] ppocr INFO: ToCHWImage : None
[2024/07/12 15:49:51] ppocr INFO: KeepKeys :
[2024/07/12 15:49:51] ppocr INFO: keep_keys : ['image', 'shape', 'polys', 'ignore_tags']
[2024/07/12 15:49:51] ppocr INFO: loader :
[2024/07/12 15:49:51] ppocr INFO: batch_size_per_card : 1
[2024/07/12 15:49:51] ppocr INFO: drop_last : False
[2024/07/12 15:49:51] ppocr INFO: num_workers : 1
[2024/07/12 15:49:51] ppocr INFO: shuffle : False
[2024/07/12 15:49:51] ppocr INFO: Global :
[2024/07/12 15:49:51] ppocr INFO: cal_metric_during_train : True
[2024/07/12 15:49:51] ppocr INFO: checkpoints : None
[2024/07/12 15:49:51] ppocr INFO: distributed : False
[2024/07/12 15:49:51] ppocr INFO: epoch_num : 500
[2024/07/12 15:49:51] ppocr INFO: eval_batch_step : [0, 2000]
[2024/07/12 15:49:51] ppocr INFO: infer_img : doc/imgs_en/img_10.jpg
[2024/07/12 15:49:51] ppocr INFO: log_smooth_window : 20
[2024/07/12 15:49:51] ppocr INFO: pretrained_model : .\pretrain_models\ch_ppocr_server_v2.0_det_train\best_accuracy
[2024/07/12 15:49:51] ppocr INFO: print_batch_step : 10
[2024/07/12 15:49:51] ppocr INFO: save_epoch_step : 3
[2024/07/12 15:49:51] ppocr INFO: save_inference_dir : None
[2024/07/12 15:49:51] ppocr INFO: save_model_dir : ./output/ch_db_res18/
[2024/07/12 15:49:51] ppocr INFO: save_res_path : ./output/det_db/predicts_db.txt
[2024/07/12 15:49:51] ppocr INFO: use_gpu : False
[2024/07/12 15:49:51] ppocr INFO: use_visualdl : False
[2024/07/12 15:49:51] ppocr INFO: Loss :
[2024/07/12 15:49:51] ppocr INFO: alpha : 5
[2024/07/12 15:49:51] ppocr INFO: balance_loss : True
[2024/07/12 15:49:51] ppocr INFO: beta : 10
[2024/07/12 15:49:51] ppocr INFO: main_loss_type : DiceLoss
[2024/07/12 15:49:51] ppocr INFO: name : DBLoss
[2024/07/12 15:49:51] ppocr INFO: ohem_ratio : 3
[2024/07/12 15:49:51] ppocr INFO: Metric :
[2024/07/12 15:49:51] ppocr INFO: main_indicator : hmean
[2024/07/12 15:49:51] ppocr INFO: name : DetMetric
[2024/07/12 15:49:51] ppocr INFO: Optimizer :
[2024/07/12 15:49:51] ppocr INFO: beta1 : 0.9
[2024/07/12 15:49:51] ppocr INFO: beta2 : 0.999
[2024/07/12 15:49:51] ppocr INFO: lr :
[2024/07/12 15:49:51] ppocr INFO: learning_rate : 0.001
[2024/07/12 15:49:51] ppocr INFO: name : Cosine
[2024/07/12 15:49:51] ppocr INFO: warmup_epoch : 2
[2024/07/12 15:49:51] ppocr INFO: name : Adam
[2024/07/12 15:49:51] ppocr INFO: regularizer :
[2024/07/12 15:49:51] ppocr INFO: factor : 0
[2024/07/12 15:49:51] ppocr INFO: name : L2
[2024/07/12 15:49:51] ppocr INFO: PostProcess :
[2024/07/12 15:49:51] ppocr INFO: box_thresh : 0.6
[2024/07/12 15:49:51] ppocr INFO: max_candidates : 1000
[2024/07/12 15:49:51] ppocr INFO: name : DBPostProcess
[2024/07/12 15:49:51] ppocr INFO: thresh : 0.3
[2024/07/12 15:49:51] ppocr INFO: unclip_ratio : 1.5
[2024/07/12 15:49:51] ppocr INFO: Train :
[2024/07/12 15:49:51] ppocr INFO: dataset :
[2024/07/12 15:49:51] ppocr INFO: data_dir : ./train_data
[2024/07/12 15:49:51] ppocr INFO: label_file_list : ['./train_data/det/train.txt']
[2024/07/12 15:49:51] ppocr INFO: name : SimpleDataSet
[2024/07/12 15:49:51] ppocr INFO: ratio_list : [1.0]
[2024/07/12 15:49:51] ppocr INFO: transforms :
[2024/07/12 15:49:51] ppocr INFO: DecodeImage :
[2024/07/12 15:49:51] ppocr INFO: channel_first : False
[2024/07/12 15:49:51] ppocr INFO: img_mode : BGR
[2024/07/12 15:49:51] ppocr INFO: DetLabelEncode : None
[2024/07/12 15:49:51] ppocr INFO: IaaAugment :
[2024/07/12 15:49:51] ppocr INFO: augmenter_args :
[2024/07/12 15:49:51] ppocr INFO: args :
[2024/07/12 15:49:51] ppocr INFO: p : 0.5
[2024/07/12 15:49:51] ppocr INFO: type : Fliplr
[2024/07/12 15:49:51] ppocr INFO: args :
[2024/07/12 15:49:51] ppocr INFO: rotate : [-10, 10]
[2024/07/12 15:49:51] ppocr INFO: type : Affine
[2024/07/12 15:49:51] ppocr INFO: args :
[2024/07/12 15:49:51] ppocr INFO: size : [0.5, 3]
[2024/07/12 15:49:51] ppocr INFO: type : Resize
[2024/07/12 15:49:51] ppocr INFO: EastRandomCropData :
[2024/07/12 15:49:51] ppocr INFO: keep_ratio : True
[2024/07/12 15:49:51] ppocr INFO: max_tries : 50
[2024/07/12 15:49:51] ppocr INFO: size : [960, 960]
[2024/07/12 15:49:51] ppocr INFO: MakeBorderMap :
[2024/07/12 15:49:51] ppocr INFO: shrink_ratio : 0.4
[2024/07/12 15:49:51] ppocr INFO: thresh_max : 0.7
[2024/07/12 15:49:51] ppocr INFO: thresh_min : 0.3
[2024/07/12 15:49:51] ppocr INFO: MakeShrinkMap :
[2024/07/12 15:49:51] ppocr INFO: min_text_size : 8
[2024/07/12 15:49:51] ppocr INFO: shrink_ratio : 0.4
[2024/07/12 15:49:51] ppocr INFO: NormalizeImage :
[2024/07/12 15:49:51] ppocr INFO: mean : [0.485, 0.456, 0.406]
[2024/07/12 15:49:51] ppocr INFO: order : hwc
[2024/07/12 15:49:51] ppocr INFO: scale : 1./255.
[2024/07/12 15:49:51] ppocr INFO: std : [0.229, 0.224, 0.225]
[2024/07/12 15:49:51] ppocr INFO: ToCHWImage : None
[2024/07/12 15:49:51] ppocr INFO: KeepKeys :
[2024/07/12 15:49:51] ppocr INFO: keep_keys : ['image', 'threshold_map', 'threshold_mask', 'shrink_map', 'shrink_mask']
[2024/07/12 15:49:51] ppocr INFO: loader :
[2024/07/12 15:49:51] ppocr INFO: batch_size_per_card : 2
[2024/07/12 15:49:51] ppocr INFO: drop_last : True
[2024/07/12 15:49:51] ppocr INFO: num_workers : 1
[2024/07/12 15:49:51] ppocr INFO: shuffle : True
[2024/07/12 15:49:51] ppocr INFO: profiler_options : None
[2024/07/12 15:49:51] ppocr INFO: train with paddle 2.6.1 and device Place(cpu)
[2024/07/12 15:49:51] ppocr INFO: Initialize indexs of datasets:['./train_data/det/train.txt']
[2024/07/12 15:49:51] ppocr INFO: Initialize indexs of datasets:['./train_data/det/val.txt']
[2024/07/12 15:49:52] ppocr INFO: train dataloader has 7 iters
[2024/07/12 15:49:52] ppocr INFO: valid dataloader has 5 iters
[2024/07/12 15:49:52] ppocr INFO: load pretrain successful from .\pretrain_models\ch_ppocr_server_v2.0_det_train\best_accuracy
[2024/07/12 15:49:52] ppocr INFO: During the training process, after the 0th iteration, an evaluation is run every 2000 iterations
Traceback (most recent call last):
File "tools/train.py", line 229, in
main(config, device, logger, vdl_writer, seed)
File "tools/train.py", line 200, in main
program.train(config, train_dataloader, valid_dataloader, device, model,
File "D:\ProgramData\PaddleOCR275\tools\program.py", line 337, in train
post_result = post_process_class(preds, batch[1])
File "D:\ProgramData\PaddleOCR275\ppocr\postprocess\db_postprocess.py", line 229, in call
src_h, src_w, ratio_h, ratio_w = shape_list[batch_index]
ValueError: too many values to unpack (expected 4)

Answered by aqingGitHub

Jul 12, 2024

谢谢！试过将Train的batch_size_per_card改成8 就没报错了

View full answer

GreatV · 2024-07-12T08:01:29Z

GreatV
Jul 12, 2024
Maintainer

换成v3、v4的配置文件试试吧

2 replies

aqingGitHub Jul 12, 2024
Author

谢谢！试过将Train的batch_size_per_card改成8 就没报错了

Answer selected by aqingGitHub

aqingGitHub Jul 17, 2024
Author

我发现把配置文件 Train 节点下的 batch_size_per_card: 设置为预训练数据train_data\det\train里的图片张数就不会报该问题了，不明所以，训练完后没看到生成best_accuracy.pdparams，一个接一个的坑啊

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

ValueError: too many values to unpack (expected 4) #13363

{{title}}

Replies: 1 comment 2 replies

{{title}}

{{title}}

{{title}}

Select a reply

ValueError: too many values to unpack (expected 4) #13363

aqingGitHub Jul 12, 2024

Replies: 1 comment · 2 replies

GreatV Jul 12, 2024 Maintainer

aqingGitHub Jul 12, 2024 Author

aqingGitHub Jul 17, 2024 Author

aqingGitHub
Jul 12, 2024

Replies: 1 comment 2 replies

GreatV
Jul 12, 2024
Maintainer

aqingGitHub Jul 12, 2024
Author

aqingGitHub Jul 17, 2024
Author