Jane comments

Results 4 comments of


                                            Jane

train on default setting on qa1 but accuracy stay about 20%

@michael-wzhu I have got the same problem. Would you please specify how did you modify the code? and what accuracy for task1 did you obtain after the modification please?

_linear() got an unexpected keyword argument 'bias_start'

In nn.py: replace `from tensorflow.python.ops.rnn_cell_impl import _linear` with `from tensorflow.contrib.rnn.python.ops.core_rnn_cell import _linear` replace `flat_out = _linear(flat_args, output_size, bias, bias_start=bias_start, scope=scope` with ` with tf.variable_scope(scope or 'Linear'): flat_out = _linear(flat_args, output_size,...

ValueError: Dimensions must be equal, but are 900 and 300 for 'model_0/prepro/prepro/fw/fw/while/basic_lstm_cell/MatMul_6' (op: 'MatMul') with input shapes: [?,900], [300,400].

@hhhhlyy @yogesh-nikose I'm encountering this problem as well. Have you solved it?

absl.flags._exceptions.UnrecognizedFlagError: Unknown command line flag 'new_emb_mat'

> flags.DEFINE_integer("new_emb_mat", 0, "new_emb_mat") He asked What value should we pass to new_emb_mat?