conv_seq2seq

conv_seq2seq copied to clipboard

tensorflow.python.framework.errors_impl.FailedPreconditionError: /; Is a directory

2

when i debug,a error have found, who can help me what's wrong,thanks tf.flags.DEFINE_string("config_paths", "\ ../example_configs/conv_seq2seq.yml,\ ../example_configs/train_seq2seq.yml,\ ../example_configs/text_metrics_bpe.yml", """Path to a YAML configuration files defining FLAG values. Multiple files can be...

jxz542189

Hi, I met this problem when i train my model using your code. Node model/conv_seq2seq/encode/conv_encoder/encoder_cnn/conv_layer_0/conv1d/Conv2D incompatible shapes: Shapes (?, 1, ?, 512) and (128, 512, 1, 19) are not compatible....

Fancy7777

With which tensorflow version does it work? I failed on r1.4 and r1.11.

xiaotaw

1. This is an farseq extension version based on the simple seq2seq framework, so i will not note problems which also happen on the original version. but one problem i...

liuliu8622

question about position embedding.

1

Thanks for you code.But I have a question about the implementation of the position embedding. It seems like position endoding is randomly initialized and updated in the training just like...

Jingyilang

Result on ReadMe ?

6

Hi @tobyyouup , I thought it would be good for people like me (who want to use your implementation) if you have result of running your implementation on the dataset....

dylanee2

The residual part in the decoder was perhaps implemented wrong?

4

This is that in [the official implementaion](https://github.com/pytorch/fairseq/blob/master/fairseq/models/fconv.py#L281): > # residual > x = (x + residual) * math.sqrt(0.5) And the following is that [in this project](https://github.com/tobyyouup/conv_seq2seq/blob/78a6e4e62a4c57a5caa9d584033a85e810fd726e/seq2seq/encoders/conv_encoder_utils.py#L183): > # add res...

eduOS

I think this is due to the version of TF, my version is r1.9 and r1.11 Two ways to solve this problem: **Method One:** 1.Emacs conv_seq2seq/example_configs/conv_seq2seq.yml 2.Remove from line 15...

flycsuu

Thanks for sharing this code. I have a question, for iwslt14, how to get the vocab.de and vocab.en files?

ahyuan

got two questions about the conv_decoder_fairseq.py

2

I have read your code and it's very help! But I have two little questions: (1) in the function **conv_block**, why reverse the second axis of the next_layer? next_layer =...

winnechan