zzw922cn/Automatic_Speech_Recognition

在用capsule network 模型出现的问题?

shanguanma opened this issue · 2 comments

在用timit数据集和capsule network 模型,具体参数如下:--mode train --level phn --model CapsuleNetwork --rnncell lstm --num_layer 2 --activation sigmoid --optimizer adam
运行出现如下错误:
tensorflow.python.framework.errors_impl.InvalidArgumentError: sequence_length(0) <= 128
[[Node: CTCLoss = CTCLoss[ctc_merge_repeated=true, ignore_longer_outputs_than_inputs=false, preprocess_collapse_repeated=false, _device="/job:localhost/replica:0/task:0/device:CPU:0"](transpose/_3, _arg_Placeholder_1_0_1, _arg_Placeholder_2_0_2, _arg_Placeholder_4_0_4)]]
@zzw922cn,麻烦您了。

iyuba commented

model.seqLengths: [128,...]

tyxr commented

在用timit数据集和capsule network 模型,具体参数如下:--mode train --level phn --model CapsuleNetwork --rnncell lstm --num_layer 2 --activation sigmoid --optimizer adam
运行出现如下错误:
tensorflow.python.framework.errors_impl.InvalidArgumentError: sequence_length(0) <= 128
[[Node: CTCLoss = CTCLoss[ctc_merge_repeated=true, ignore_longer_outputs_than_inputs=false, preprocess_collapse_repeated=false, _device="/job:localhost/replica:0/task:0/device:CPU:0"](transpose/_3, _arg_Placeholder_1_0_1, _arg_Placeholder_2_0_2, _arg_Placeholder_4_0_4)]]
@zzw922cn,麻烦您了。

解决了吗?我遇到了同样的问题