Comments (3)
@RobertXiaobinW 您好,需要提供更详细的信息,最好是能够复现问题粒度
from crnn.pytorch.
这个是loss:tensor(inf, device='cuda:0', grad_fn = '<' MeanBackward0 '>' )
这个是input_length: tensor([1, 1, 1, 1, 1, 1, 1, 1])
这个是target_length: tensor([18, 9, 17, 6, 20, 11, 15, 13])
这个是target:
tensor([[ 788, 12398, 16864, 10277, 27893, 26725, 11970, 12494, 24441, 12486,
651, 25194, 21985, 10739, 28449, 18115, 24168, 24136, 0, 0,
0, 0, 0, 0, 0, 0],
[18490, 13217, 26361, 8336, 13865, 21787, 8350, 9127, 14206, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0],
[21562, 11474, 8698, 8510, 26653, 98, 9792, 26585, 18390, 9982,
12594, 27589, 15995, 1506, 25033, 29164, 27935, 0, 0, 0,
0, 0, 0, 0, 0, 0],
[22313, 20842, 22277, 28439, 26443, 9748, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0],
[30493, 20836, 25265, 23656, 24147, 15982, 27451, 28809, 28860, 11399,
16812, 8996, 17091, 27334, 10614, 16839, 28066, 15049, 21229, 21864,
0, 0, 0, 0, 0, 0],
[14392, 28384, 12272, 22050, 13004, 16370, 19666, 8364, 47, 18037,
15336, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0],
[25636, 16725, 755, 24788, 11787, 17582, 20783, 26430, 24712, 8278,
25356, 13494, 13120, 29164, 9054, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0],
[24467, 10223, 13797, 9870, 17309, 28066, 14946, 14523, 10659, 8469,
27858, 9383, 19832, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0]])
是CTC方面出了问题么?
from crnn.pytorch.
我从0开始训练也是inf,怎么解决
from crnn.pytorch.
Related Issues (20)
- 请教一下如何训练自己的数据集,数据格式以及存放目录怎么配置 HOT 2
- 问下,你验证集上,acc可以大概达到多少 HOT 4
- 用自己的字典替换你的all_word.txt文件,训练报错 HOT 11
- 下载目录无权限 HOT 2
- hi,thank you for the great codes. But I have met some problems. I just download the code, and run using distribute training mode, HOT 7
- 训练数据怎么生成的? HOT 11
- 请教 HOT 15
- 请教*2 HOT 2
- training data generation HOT 3
- result output confidence
- 单卡和多卡设置问题 HOT 1
- 印刷体纯数字识别效果不是太好,有解决办法吗?
- number of dims don't match in permute HOT 1
- 是否可以修改预训练模型的输出层后进行迁移学习呢?
- 这个不可以转ONNX测试吧
- 我尝试用cpu训练。报错了,怎么解决?微信nlanguage 。 py -3 train.py --direction horizontal
- 我在all words.txt里放了10个汉字,尝试运行generator.py,报了以下错误,怎么解决?微信nlanguage Traceback (most recent call last): File "F:/pycharm2020.2/crnn.pytorch_generator/generator.py", line 227, in <module> test_image_gen('horizontal') File "F:/pycharm2020.2/crnn.pytorch_generator/generator.py", line 207, in test_image_gen im, indices, target_len = gen.gen_image() File "F:/pycharm2020.2/crnn.pytorch_generator/generator.py", line 158, in gen_image text = np.random.choice(FONT_CHARS_DICT[font_path], target_len) File "mtrand.pyx", line 908, in numpy.random.mtrand.RandomState.choice ValueError: 'a' cannot be empty unless no samples are taken
- win10直接运行train.py 报错。微信nlanguage ForkingPickler(file, protocol).dump(obj) TypeError: cannot pickle 'Font' object HOT 1
- 字符集中有的字符例如\u3000在python下没法显示报错,怎么去掉?微信nlanguage
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from crnn.pytorch.