This tag is used for Google's deprecated seq2seq framework, an encoder-decoder framework for Tensorflow (revamped version is called Neural Machine Translation)
Questions tagged [sequence-to-sequence]
94 questions
1
vote
0 answers
Why is Paddle throwing errors when feeding in a dense_vector_sequence to a SeqToseq model?
I am trying to use paddle-paddle (https://github.com/baidu/Paddle) to train a (encoder-decoder) sequence to sequence model for POS tagging.
But instead of using a one-hot embedding of the word indices as input, I would be using an imaginary word…
alvas
- 115,346
- 109
- 446
- 738
0
votes
0 answers
How to perform inference for sequence 2 sequence models in tensorflow?
I am trying to build a sequence-to-sequence model in TensorFlow from scratch for English to German translation. After training the model when I try to do inference using a sentence it doesn't predict anything else other than token. I just need a…
Na462
- 11
- 2
0
votes
0 answers
The output sequence is almost uniform at each time step when setting return_sequence=True
I'm using LSTM (python/tensorflow) to build a model. We have a time sequence of the input X_train in shape of (, 8). The output is a single value. The model performs very well by setting return_sequence=False.
lstm_model =…
0
votes
0 answers
NLP for Sign Language
I have successfully developed an ML model that effectively recognizes American Sign Language using the pre-trained "Gesture Recognizer" model from MediaPipe (https://developers.google.com/mediapipe/solutions/vision/gesture_recognizer).
However, I am…
Ahmed Adel
- 49
- 6
0
votes
0 answers
Time series prediction problem for seq2seq data
I have a sequence prediction problem that confuses me. I am trying to predict some graphs as an output. I am using LSTM for this. I tried several things but still there is not any good results. I have 158 laboratory test which has 4 time series as…
omer_tb
- 11
- 2
0
votes
0 answers
Inferencing in encoder-decoder teacher forcing model with floating point values
I am trying to model a translation between two numerical (floating point) datasets and thought of using sequence to sequence learning with teaching enforcement. I am able to run the training model with a decently low mse but when it comes to the…
0
votes
0 answers
AttributeError: module 'tensorflow' has no attribute 'contrib' . How do I overcome this?
I have codes with remanants of Tensorflow v1.0 , but i want to adapt it to 2.0 This is the bit of code where it keeps stopping ? how do i adapt this ?
# Building the seq2seq model
def seq2seq_model(inputs, targets, keep_prob, batch_size,…
0
votes
0 answers
predicting output symbols in parallel in the seq-to-seq task
Though I am not sure if the stackoverflow is a right place to ask a theoritical question which is not directly related to a programming topic, I will just post my question.
One of motivations in developing the Transformer was to achieve parallelism…
MLER
- 26
- 3
0
votes
1 answer
Workaround / fallback value for tfp.distributions.Categorical.log_prob in tensorflow graph mode
Is there a way to avoid tfp.distributions.Categorical.log_probraising an error if the input is a label out of range?
I am passing a batch of samples to the log_prob method, some of them have the value n_categories + 1, which is what you get as…
RR_28023
- 158
- 2
- 12
0
votes
1 answer
Many-to-Many LSTM PyTorch
I want to build an LSTM model for the FashionMNIST dataset in PyTorch. I will later on need to extend this to a different dataset that contains videos.
It should get a sequence of images (of the FashionMNIST) as the input (let's say 20 images) and…
Jam
- 1
- 2
0
votes
1 answer
How do I decode the output of my seq-to-seq model if I'm using an embedding layer?
I have a seq to seq model trained of some clever bot data:
justphrases_X is a list of sentences and justphrases_Y is a list of responses to those sentences.
maxlen = 62
#low is a list of all the unique words.
def…
Haztec
- 31
- 3
0
votes
1 answer
many to many sequence prediction variable length input/output inkeras
Im trying to to predict a variable length input/output many to many sequence using Keras, the dataframe below is a representation of the data . 5 columns and one target column.
df3={'email':…
sr33kant
- 35
- 3
0
votes
2 answers
AssertionError: Could not compute output Tensor("softmax_layer/Identity:0", shape=(None, 27, 8870), dtype=float32)
I am trying to develop chatbot with an attention mechanism. but it gives errors like this. my input shape of x_train is (None, 27) and output shape is (None, 27, 8870). But I can't identify the errors properly.
def chatbot_model(embedding_size,…
Vivek Dhinoja
- 13
- 4
0
votes
1 answer
Training in inference mode in seq-to-seq model
This is apparently the code for seq2seq model with embedding that i wrote
encoder_inputs = Input(shape=(MAX_LEN, ), dtype='int32',)
encoder_embedding = embed_layer(encoder_inputs)
encoder_LSTM = LSTM(HIDDEN_DIM, return_state=True)
…
saransh bhatnagar
- 40
- 4
0
votes
1 answer
Tensorflow in C++: Successfully reading the protocol buffer for LSTM encoder decoder model
I have been debugging this issue for a while now. I have developed an LSTM encoder decoder model which I plan to deploy in C++.
Having saved the model in the .pb file format, I am able to import the model and data and deploy it within python.…
Narayan Narvekar
- 1
- 1