我运行multigpu_train.py的时候出现问题

[songmimi]

Generator use 10 batches for buffering, this may take a while, you can tune this yourself.
11 training images in data/ocr/icdar
text file gt_data/ocr/icdar/img_994.txt does not exists
text file gt_data/ocr/icdar/img_993.txt does not exists
text file gt_data/ocr/icdar/img_999.txt does not exists
text file gt_data/ocr/icdar/img_991.txt does not exists
text file gt_data/ocr/icdar/img_995.txt does not exists
这是我的代码运行结果，老是显示gt格式不存在，可是我的gt在的，而且我的icdar2015数据集是我在官网下载的，我的问题困扰我好久了，希望大家帮我一下

[WoshiBoluo]

亲你现在能正常跑起来了吗

[ho1messi]

readme 里面 Train 部分有提到要改标注文件名称或者代码
改代码的话就把 icdar.py 里 604 行改成
txt_fn = im_fn.replace(os.path.basename(im_fn), 'gt_%s.txt' % os.path.basename(im_fn).split('.')[0])
就好了

Note: you should change the gt text file of icdar2015's filename to img_.txt instead of gt_img_.txt(or you can change the code in icdar.py), and some extra characters should be removed from the file. See the examples in training_samples/

[pankSM]

您好,请问您训练时候怎么使用的的gpu?我按照教程来的,结果只使用显卡60M,
基本没使用gpu