shawn
shawn
for example, the task is to classify dog or not, and we label dog as class 1. when comes a dog image, class 1 focus on any part except dog;...
I have several hundred GB wav files on my disk (about 1, 000 hours wav data). I found directly reading the wav file is slow for training, so I choose...
python 3.6 latest ChineseTone version ` ChineseTone.PinyinHelper.hasMultiPinyin('鮟') True ChineseTone.PinyinHelper.hasMultiPinyin('鱇') True ChineseTone.PinyinHelper.hasMultiPinyin('徐') True ` in addition, characters like '检、海' etc. are all recognized as poly-phone character.
in your original code `opt = SGD(trainable_params(model), momentum=0.9, weight_decay=5e-4*batch_size, nesterov=True)` multiply 5e-4 by batch_size is redundant.
@zyzisyz any plan to test this dataset?