ZhangZhaofeng

ZeroDivisionError: float division by zero in loss function

5

Run script in > dense_correspondence/experiments/shoes/training_shoes.ipynb meet a ZeroDivisionError Dataset is > shoe_train_4_shoes.yaml It seems that in some case, the num_matches == 0 cause the error. How to fix it? Error...

Can this code work on a tensorflow trained model

6

Hi Kumar I see in steps_kt/decode_seq.sh ,line of 75 `export KERAS_BACKEND=theano.` It seems that this code can only run on theano backend. But in my case, the model is trained...

getBinaryLabels function not defined

1

File "/tools/ASR/steps_kt/dataGenSequences.py", line 134, in getNextSplitData labelMat = self.getBinaryLabels(labels[uid]) AttributeError: 'dataGenSequences' object has no attribute 'getBinaryLabels'

灵活选择gpu

2

在infer阶段，如果有两个以上的gpu，程序会默认选择所有gpu进行推理。 这时如果有gpu已经被占用，就会发生内存不足的错误。 希望在infer阶段加入选择gpu的选项。

srp一般是多通道的gcc相加，这里只看见2通道的实现。希望能看到更新多通道的情况

https://github.com/orchidas/Speech-Denoising/blob/f2b227d2a4f04dc8710e9ec0cdf1549ea39312bd/cmi_tuned_kalman_speech.m#L20 this line should be totseg=ceil((length(y)-overlap)/(l-overlap)); The original line would in some cases count an extra frame, which would lead to an array overflow error.