## Train CRNN
To train your own model, firstly you need to prepare your own text-line dataset, and organize your dataset with an infoile.
In the infofile, each line represents an text-line image, path to image and image label are splited by a special character '\t'.
Such as follows:
>data_set/my_data1/0001.jpg\t37918
data_set/my_data1/0002.jpg\tHello World!
data_set/my_data1/0003.jpg\t你好
...
Then replace your infofile by your own in the training file train_warp_ctc.py
>config.train_infofile = ['path_to_train_infofile1.txt','path_to_train_infofile2.txt']
config.val_infofile = 'path_to_test_infofile.txt'
Then if you want to use warp-ctc as loss function, run
>python3 train_warp_ctc.py
or if you want to use pytorch-ctc as loss function, run
>python3 train_pytorch_ctc.py
python库pytorch学习资料
需积分: 9 88 浏览量
2022-10-19
22:30:22
上传
评论
收藏 522.34MB ZIP 举报
python库pytorch学习资料 （2000个子文件） 
TextRNN.ckpt 8.65MB TextCNN.ckpt 8.12MB model.ckpt 0B temps.csv 11KB Dockerfile 579B
horse2zebra.gif 7.33MB
mnist.pkl.gz 15.42MB train-images-idx3-ubyte.gz 9.45MB train-images-idx3-ubyte.gz 9.45MB t10k-images-idx3-ubyte.gz 1.57MB t10k-images-idx3-ubyte.gz 1.57MB train-labels-idx1-ubyte.gz 28KB train-labels-idx1-ubyte.gz 28KB t10k-labels-idx1-ubyte.gz 4KB t10k-labels-idx1-ubyte.gz 4KB
LSTM.html 340KB 图像识别实战常用模块解读.ipynb 2.42MB ═╝╧±╩╢▒≡╩╡╒╜│ú╙├─ú┐Θ╜Γ╢┴-checkpoint.ipynb 2.42MB Flowers_Classification-checkpoint.ipynb 2.42MB ╛φ╗²═°┬τ╩╡╒╜-checkpoint.ipynb 800KB 2 Gradient Descent-checkpoint.ipynb 294KB 搭建PyTorch神经网络进行气温预测.ipynb 195KB ┤ε╜¿PyTorch╔±╛¡═°┬τ╜°╨╨╞°╬┬╘ñ▓Γ-checkpoint.ipynb 134KB 4 Multiple Linear Regression-checkpoint.ipynb 69KB nn_tutorial.ipynb 44KB 强大的hub模块.ipynb 41KB 神经网络分类任务.ipynb 24KB ╔±╛¡═°┬τ╖╓└α╚╬╬±-checkpoint.ipynb 23KB autograd机制.ipynb 15KB ╝╕╓╓╨╬╫┤╡─Tensor-checkpoint.ipynb 14KB 卷积神经网络效果.ipynb 11KB ╛φ╗²╔±╛¡═°┬τ╨º╣√-checkpoint.ipynb 11KB PyTorch简介与安装.ipynb 9KB 1-PyTorch╝≥╜Θ╙δ░▓╫░-checkpoint.ipynb 9KB CycleGAN.ipynb 8KB pix2pix.ipynb 7KB autograd╗·╓╞-checkpoint.ipynb 7KB 几种形状的Tensor.ipynb 6KB 几种形状的Tensor-checkpoint.ipynb 6KB word2vec.ipynb 5KB ╟┐┤≤╡─hub─ú┐Θ-checkpoint.ipynb 72B word2vec-checkpoint.ipynb 72B
img_7100.jpg 1.63MB img_7023.jpg 1.5MB img_7048.jpg 1.47MB img_7039.jpg 1.35MB img_7050.jpg 1.29MB img_7009.jpg 1.29MB img_7049.jpg 1.28MB img_7018.jpg 1.24MB img_7011.jpg 1.24MB img_7052.jpg 1.22MB img_7114.jpg 1.21MB img_7002.jpg 1.2MB img_7058.jpg 1.2MB img_7015.jpg 1.18MB img_7063.jpg 1.18MB img_7066.jpg 1.17MB img_7077.jpg 1.15MB img_7056.jpg 1.14MB img_7121.jpg 1.14MB img_7022.jpg 1.14MB img_7091.jpg 1.13MB img_7059.jpg 1.13MB img_7040.jpg 1.13MB img_7089.jpg 1.12MB img_7065.jpg 1.12MB img_7045.jpg 1.12MB img_7025.jpg 1.12MB img_7055.jpg 1.11MB img_7035.jpg 1.11MB img_7053.jpg 1.11MB img_7051.jpg 1.11MB img_7041.jpg 1.1MB img_7135.jpg 1.1MB img_7072.jpg 1.1MB img_7034.jpg 1.1MB img_7088.jpg 1.1MB img_7016.jpg 1.1MB img_7067.jpg 1.1MB img_7090.jpg 1.1MB img_7130.jpg 1.09MB img_7064.jpg 1.09MB img_7076.jpg 1.09MB img_7082.jpg 1.09MB img_7029.jpg 1.09MB img_7062.jpg 1.09MB img_7069.jpg 1.09MB img_7070.jpg 1.09MB img_7017.jpg 1.09MB img_7044.jpg 1.09MB img_7010.jpg 1.08MB img_7068.jpg 1.08MB img_7043.jpg 1.08MB img_7092.jpg 1.08MB img_7008.jpg 1.08MB img_7042.jpg 1.07MB img_7095.jpg 1.07MB img_7087.jpg 1.07MB img_7027.jpg 1.07MB共 2000 条
- 1
- 2
- 3
- 4
- 5
- 6
- 20
评论0