Hong
Results
3
comments of
Hong
# It might be the problem of params here. Set max_seq_len a larger value is ok for me! class PositionalEncoder(nn.Module): def __init__(self, d_model, max_seq_len = **200**, dropout = 0.1): super().__init__()...
### change the code in beam.py as following. It seems there is a bug in the origin code. ### the origin code try to stop when there are k eos...
Is the data available now? Thank you very much!