请教一下机器学习PAI，我的时间差序列改成以下这个配置，请问报错怎么解决？
features: {
input_names: ‘pc_log_times_diff_ss’
feature_type: SequenceFeature
sub_feature_type: RawFeature
separator: ‘!’
embedding_dim: 8
sequence_combiner{
attention{}
}
}
报错logview:
http://logview.alibaba-inc.com/logview/?h=http://service.odps.aliyun-inc.com/api&p=b_risk_dev&i=20231115021909796glars1gvyg2_cac0bd82_aa46_4546_b683_8a364101a786&token=RC9zTjk0eWQzdUlINHNIU3J6VFlaUWdoVHBNPSxPRFBTX09CTzoxNDk2MzI3NTcyMDcyNzY0LDE3MDI2MDY3NTQseyJTdGF0ZW1lbnQiOlt7IkFjdGlvbiI6WyJvZHBzOlJlYWQiXSwiRWZmZWN0IjoiQWxsb3ciLCJSZXNvdXJjZSI6WyJhY3M6b2RwczoqOnByb2plY3RzL2Jfcmlza19kZXYvaW5zdGFuY2VzLzIwMjMxMTE1MDIxOTA5Nzk2Z2xhcnMxZ3Z5ZzJfY2FjMGJkODJfYWE0Nl80NTQ2X2I2ODNfOGEzNjQxMDFhNzg2Il19XSwiVmVyc2lvbiI6IjEifQ==
看 stderr 的报错：
raise type(e)(node_def, op, message)
tensorflow.python.framework.errors_impl.InvalidArgumentError: StringToNumberOp could not correctly convert string:
[[{{node sequence_str_2_float_pc_log_times_diff_ss}} = StringToNumberout_type=DT_FLOAT, _device=”/device:CPU:0″]]
里面说你定义的是 float, 但是其实是 string, 没法转成 float
排查方法：