iamweiweishi
iamweiweishi
FormatMultiTaskLabels Where can I find the above class? Thank you.
Thank you so much. @usimarit I am trying to retrain a stream-model with params: speech_config: sample_rate: 16000 frame_ms: 25 stride_ms: 10 num_feature_bins: 80 feature_type: log_mel_spectrogram preemphasis: 0.97 normalize_signal: False normalize_feature:...
@usimarit I trained one new stream-transducer model with the modified params above. Then, I tried to infer with the scipt below referred to [this](https://github.com/TensorSpeech/TensorFlowASR/blob/6247a5c86bcaacf87b20345533d65af85b1894d6/tests/featurizer/test_speech_featurizer.ipynb), but still could not get the...
- k = 20 failed failed failed shui shui di shui di di yi shui di di yi yan yan shui di di yi yan yan yi shui di di...
同样的错误，请问下 有没有解决的方法？
Hi, I trained one model using res50. And I tried to convert the trained models to binarized ones. Unfortunately, never succeed. I found the parameter 'alphas_' in 'binary_conv_layer.cpp', but I...
from six.moves import input, zip as izip
I found that the MXNet should be recompiled again and then add the multi_acc function in the light_moon.py.
official MXNet
You should copy the files in src/operator into MXNet and recompile MXNet again.