I’m a Backend - AI- Blockchain engineer in developing and designing distributed applications.
Professional skills:
activation_75 (Activation) (None, 3, 18, 384) 0 batch_normalization_75[0][0]
max_pooling2d_2 (MaxPooling2D) (None, 3, 18, 320) 0 block35_10_ac[0][0]
Total params: 4,342,240
Trainable params: 4,333,024
Non-trainable params: 9,216
data train: 11214
Traceback (most recent call last):
File "run.py", line 61, in
total_loss += train_step(imgs, target)
File "run.py", line 32, in train_step
loss += loss_function(word_target[:, i, :], y_pred)
File "C:\Users\1619872\OneDrive - Standard Chartered Bank\Desktop\attention_OCR_master1\attention-ocr-master\metrics.py", line 16, in loss_function
loss = tf.keras.losses.categorical_crossentropy(y_true, y_pred)
File "C:\Users\1619872\python-env\local\lib\site-packages\tensorflow\python\util\dispatch.py", line 206, in wrapper
return target(*args, **kwargs)
File "C:\Users\1619872\python-env\local\lib\site-packages\keras\losses.py", line 1666, in categorical_crossentropy
y_true, y_pred, from_logits=from_logits, axis=axis)
File "C:\Users\1619872\python-env\local\lib\site-packages\tensorflow\python\util\dispatch.py", line 206, in wrapper
return target(*args, **kwargs)
File "C:\Users\1619872\python-env\local\lib\site-packages\keras\backend.py", line 4839, in categorical_crossentropy
target.shape.assert_is_compatible_with(output.shape)
File "C:\Users\1619872\python-env\local\lib\site-packages\tensorflow\python\framework\tensor_shape.py", line 1161, in assert_is_compatible_with
raise ValueError("Shapes %s and %s are incompatible" % (self, other))
ValueError: Shapes (64, 163) and (64, 160) are incompatible
What dataset did you used for Attention_OCR training? Original paper uses tfrecord files, but this project uses simpler data format.
Hi!After four or five days of training, 60000 verification codes synthesized by ourselves can not accurately identify the verification codes. What may be the reason for this
Hello ông, tôi có làm và huấn luyện theo code của ông. Sau khi huấn luyện xong mô hình tôi tiến hành predict, mọi kết quả với ảnh đầu vào là ảnh trong tập dữ liệu train đều cho ra kết quả chuẩn. Tuy nhiên khi cho vào những ảnh khác không xuất hiện trong tập train thì kết quả hoàn toàn không chính xác. Fomat dữ liệu nó cho ra là chuẩn, chẳng hạn như 30-B5 21187, ... y như trong file txt mình gán nhãn. Tuy nhiên các giá trị lại hoàn toàn không chính xác. hi vọng ông có thể giúp tôi tìm hiểu thêm.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.