I increased the VOCAB_SIZE to prevent unk in the answers, however, I was faced with the problem of lack of memory. Could you tell me please, what parameters can be change to avoid this?

File "main_simple_seq2seq.py", line 18, in
from data.twitter import data
ImportError: No module named twitter

Did I miss some files？
Can you please help me？Many thanks^ o^

Can someone share with me how to resolve this error? Thanks.

Traceback (most recent call last):
File "D:\ChatBot\seq2seq-chatbot-master\main.py", line 11, in
from tensorlayer.models.seq2seq import Seq2seq
ModuleNotFoundError: No module named 'tensorlayer.models.seq2seq'

Hi, I was wondering how I could train my own data, is there a specific format I need my data in?

python3 main.py

Traceback (most recent call last):
File "main.py", line 12, in
from tensorlayer.models.seq2seq_with_attention import Seq2seq_Attention

i have this error: Load n.npz failed!
May I know why it is like this?

After a few hours of debugging, uninstalling, reinstalling, and generally trying to get this code working, I've found the root cause of the problem:

1. This repository requires TensorFlow 2.0.0+

2. This repository also requires Python 3.7+

3. TensorFlow 2.0.0+ only supports Python 3.6 and below.

Can someone provide some guidance on how to get TensorFlow working on a version of Python that isn't supported?

Hello, I'm planning to train this with my GPU to make the training faster. The problem is it only utilizes my CPU. Thank you.

So I can continue train with my own data

Is there any code to use data of cornell_corpus?

Hello, I have a problem when I load my trained model ("model.npz"). It gives me this error:
ValueError: Dimension 0 in both shapes must be equal, but are 6004 and 8004. Shapes are [6004,1024] and [8004,1024]. for 'Assign' (op: 'Assign') with input shapes: [6004,1024], [8004,1024].

Could anyone explain to me a little bit? Or do I need to modify the code?

Many thanks! :)

Training on CPU is taking too much time, so do you have any estimate how much time it will take? I have executed this 12 hours ago and now i am on just "Epoch[2/50] step:[600/2852] loss:5.684645 took:9.62770s". Can you please help me to boost this training.

hi
is there a pretrained model download? i tried to train a model but it was to slow with my cpu .
@tensorlayer-gardener

TypeError: () got an unexpected keyword argument 'is_last'

In Practicle seq2seq, it provides a checkpoint file, but in tensorlayer 2.x.x, there was no load_ckpt anymore, so how to convert it into .npz file

When I run the program, there are lots of [TL]*** messages generated. Is there a way to hide these msg? It seems not much online information about this. Thanks in advance!

Dear Dr.pskrunner

Hello, i am a follower of your github. i have a question that whether the copy mechanism (copy_decoder)is available in pskrunner14/pytorch-seq2seq or not？

Because i want to learn that code.

Thank you, very much.

Best wishes

Tianlin Zhang

flake8 testing of https://github.com/zsdonghao/seq2seq-chatbot on Python 2.7.13

$ flake8 . --count --select=E901,E999,F821,F822,F823 --show-source --statistics

./tensorlayer/layers.py:5976:49: F821 undefined name 'single_cell'
            cell = tf.contrib.rnn.MultiRNNCell([single_cell] * num_layers)
                                                ^

./tensorlayer/layers.py:5978:49: F821 undefined name 'single_cell'
            cell = tf.nn.rnn_cell.MultiRNNCell([single_cell] * num_layers)
                                                ^

./tensorlayer/prepro.py:1007:8: F821 undefined name 'is_random'
    if is_random:
       ^

I am using the seq2seq model as an autoencoder. Given a test paragraph, I'd like to get the thought vector (using the terminology in the figure of README.md).

ModuleNotFoundError Traceback (most recent call last)
in ()
8
9 ###============= prepare data
---> 10 from data.twitter import data
11 metadata, idx_q, idx_a = data.load_data(PATH='data/twitter/') # Twitter
12 # from data.cornell_corpus import data

ModuleNotFoundError: No module named 'data.twitter'

I am trying to write code for Chat Box, but encountering the error "ModuleNotFoundError: No module named 'data.twitter'; 'data' is not a package" when trying to execute "from data.twitter import data".

Please suggest , how to resolve the issue?

note: I am working on following environment:
Python is 3.6 V
Tensorflow : 2.0
Tensorlayer: 2.2
python-twitter

How any I chane the dataset . Can I take the dataset of mine rather then twitter data & corpus movie data
if yes then how?

Exception: Layer 'model/embedding/seq_embedding' already exists, please choice other 'name' or reuse this layer
Hint : Use different name for different 'Layer' (The name is used to control parameter sharing)
Additional Informations: http://tensorlayer.readthedocs.io/en/latest/modules/layers.html?highlight=clear_layers_name#tensorlayer.layers.clear_layers_name

when i run the code, this problem comes.

python: 3.6
tensorflow-gpu: 1.7.1
tensorlayer: 1.7.4

Hi,

It seems that the example didn't use Word2Vec. I am trying to load pre-trained vectors to represent these words. Do you think it will be helpful to to improve the bot performance? I can't figure out how to do that. Do you have any examples showing the similar procedures?

Thank you in advance!

python main_simple_seq2seq.py
Traceback (most recent call last):
File "main_simple_seq2seq.py", line 18, in
from data.twitter import data
ImportError: No module named twitter

I am trying to serve the model over tensorflow serving and I have created the below signature. But it doesnt seem to work. Please help me @pskrunner14

encode_seqs = tf.placeholder(dtype=tf.int64, shape=[batch_size, None], name="encode_seqs")
decode_seqs = tf.placeholder(dtype=tf.int64, shape=[batch_size, None], name="decode_seqs")

Inference Data Placeholders

encode_seqs2 = tf.placeholder(dtype=tf.int64, shape=[1, None], name="encode_seqs")
decode_seqs2 = tf.placeholder(dtype=tf.int64, shape=[1, None], name="decode_seqs")

export_path_base = './export_base/'
export_path = os.path.join(
tf.compat.as_bytes(export_path_base),
tf.compat.as_bytes(str(1)))
print('Exporting trained model to', export_path)
builder = tf.saved_model.builder.SavedModelBuilder(export_path)

    classification_inputs = tf.saved_model.utils.build_tensor_info(
        encode_seqs)
    classification_outputs_classes = tf.saved_model.utils.build_tensor_info(
        decode_seqs)
    #classification_outputs_scores = tf.saved_model.utils.build_tensor_info(loss)

    classification_signature = (
        tf.saved_model.signature_def_utils.build_signature_def(
            inputs={
                tf.saved_model.signature_constants.CLASSIFY_INPUTS:
                    classification_inputs
            },
            outputs={
                tf.saved_model.signature_constants.CLASSIFY_OUTPUT_CLASSES:
                    classification_outputs_classes,
                #tf.saved_model.signature_constants.CLASSIFY_OUTPUT_SCORES:
                    #classification_outputs_scores
            },
            method_name=tf.saved_model.signature_constants.CLASSIFY_METHOD_NAME))

    tensor_info_x = tf.saved_model.utils.build_tensor_info(encode_seqs2)
    tensor_info_y = tf.saved_model.utils.build_tensor_info(decode_seqs2)

    prediction_signature = (
        tf.saved_model.signature_def_utils.build_signature_def(
            inputs={'issue': tensor_info_x},
            outputs={'solution': tensor_info_y},
            method_name=tf.saved_model.signature_constants.PREDICT_METHOD_NAME))

    builder.add_meta_graph_and_variables(
        sess, [tf.saved_model.tag_constants.SERVING],
        signature_def_map={
            'predict_solution':
                prediction_signature,
            tf.saved_model.signature_constants.DEFAULT_SERVING_SIGNATURE_DEF_KEY:
                classification_signature,
        },
        main_op=tf.tables_initializer(),
        strip_default_attrs=True)

    builder.save()

    print('Done exporting!')

I have the below signature,

C:\Users\d074437\PycharmProjects\seq2seq>saved_model_cli show --dir ./export_base/1 --all

MetaGraphDef with tag-set: 'serve' contains the following SignatureDefs:

signature_def['predict_solution']:
The given SavedModel SignatureDef contains the following input(s):
inputs['issue'] tensor_info:
dtype: DT_INT64
shape: (1, -1)
name: encode_seqs_1:0
The given SavedModel SignatureDef contains the following output(s):
outputs['solution'] tensor_info:
dtype: DT_INT64
shape: (1, -1)
name: decode_seqs_1:0
Method name is: tensorflow/serving/predict

signature_def['serving_default']:
The given SavedModel SignatureDef contains the following input(s):
inputs['inputs'] tensor_info:
dtype: DT_INT64
shape: (32, -1)
name: encode_seqs:0
The given SavedModel SignatureDef contains the following output(s):
outputs['classes'] tensor_info:
dtype: DT_INT64
shape: (32, -1)
name: decode_seqs:0
Method name is: tensorflow/serving/classify

But when I try to run it, I get an error as below

C:\Users\d074437\PycharmProjects\seq2seq>saved_model_cli run --dir ./export_base --tag_set serve --signature_def predict_solution --inputs='this is the text'
usage: saved_model_cli [-h] [-v] {show,run,scan} ...
saved_model_cli: error: unrecognized arguments: is the text'

I keep getting this error every time validation is going on. and also for testing.

Could you explain me how to solve it ?

Hi you! I don't see you use validX,validY data in your code. Could you tell me how to get loss of valid data in your code. Thanks you!

validx and validy are not used and there are no calculcations for validation accuracy in the model ?

After training, I manage to get some answers, but most often instead of words I get a strange "unk" in the answer.

Hi there,

I trained the data for a few days and now the samples are returning good results to the predefined "Happy Birthday" and "Trump" requests.

Great job by you. Thanks so far.

Do you already have a small python program for using the chatbot? If I write a message, the chatbot should return a single answer.

Thanks
Chris

I posted this under a work account so I closed the issue and logged in with my personal.

The issue I get is this:

python3 main.py

Traceback (most recent call last):
File "main.py", line 11, in
from tensorlayer.models.seq2seq import Seq2seq
ImportError: No module named 'tensorlayer.models.seq2seq'

i say hi, it sometimes says random stuff, sometimes replies hi darling. Can anyone please help me understand what's happening

I have trained the model on the dataset but I don't know how to get the output using .npz file. Please add instructions to test it as well.

When I run the main script this warning message shows up. Any ideas how to rectify this?

[TL] WARNING: From main_simple_seq2seq.py:144: initialize_global_variables (from tensorlayer.layers.utils) is deprecated and will be removed after 2018-09-30. Instructions for updating: This API is deprecated in favor of tf.global_variables_initializer

I tried replacing it with tf.global_variables_initializer but the script crashed with the following message:

FailedPreconditionError (see above for traceback): Attempting to use uninitialized value model/seq2seq/encode/rnn/multi_rnn_cell/cell_2/basic_lstm_cell/kernel [[Node: model/seq2seq/encode/rnn/multi_rnn_cell/cell_2/basic_lstm_cell/kernel/read = Identity[T=DT_FLOAT, _device="/job:localhost/replica:0/task:0/device:GPU:0"](model/seq2seq/encode/rnn/multi_rnn_cell/cell_2/basic_lstm_cell/kernel)]] [[Node: seq_loss_with_mask/_51 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/device:CPU:0", send_device="/job:localhost/replica:0/task:0/device:GPU:0", send_device_incarnation=1, tensor_name="edge_2513_seq_loss_with_mask", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/device:CPU:0"]()]]

Would you please advise?