Comments (3)
EDIT: Added an example to use RNNs with variable seq length in TensorFlow: Here
Hi, for me their example is just working fine! You can provide different values for early_stop, the timestep you want to stop at. Maybe make sure you are using tensorflow 0.6.0.
import tensorflow as tf
from tensorflow.models.rnn import rnn
from tensorflow.models.rnn.rnn_cell import BasicLSTMCell, LSTMCell
import numpy as np
if __name__ == '__main__':
np.random.seed(1)
size = 100
batch_size= 100
n_steps = 45
seq_width = 50
initializer = tf.random_uniform_initializer(-1,1)
seq_input = tf.placeholder(tf.float32, [n_steps, batch_size, seq_width])
#sequence we will provide at runtime
early_stop = tf.placeholder(tf.int32)
#what timestep we want to stop at
inputs = [tf.reshape(i, (batch_size, seq_width)) for i in tf.split(0, n_steps, seq_input)]
#inputs for rnn needs to be a list, each item being a timestep.
#we need to split our input into each timestep, and reshape it because split keeps dims by default
cell = LSTMCell(size, seq_width, initializer=initializer)
initial_state = cell.zero_state(batch_size, tf.float32)
outputs, states = rnn.rnn(cell, inputs, initial_state=initial_state, sequence_length=early_stop)
#set up lstm
iop = tf.initialize_all_variables()
#create initialize op, this needs to be run by the session!
session = tf.Session()
session.run(iop)
for e_s in [10, 100, 200, 250]:
feed = {early_stop:e_s, seq_input:np.random.rand(n_steps, batch_size, seq_width).astype('float32')}
outs = session.run(outputs, feed_dict=feed)
print len(outs)
from tensorflow-examples.
Have a look at https://danijar.com/variable-sequence-lengths-in-tensorflow/
from tensorflow-examples.
Hi! For a faster implementation using CuDNN, check this out: tensorflow/tensorflow#22308
from tensorflow-examples.
Related Issues (20)
- 404 not
- 404 not found
- Please provide a example for stacked bidirectional LSTM for Tensorflow 2.x
- [Potential NAN bug] Loss may become NAN during training HOT 1
- Tensor
- How to get prediction code ?
- fixes for Word2Vec for Python 3
- ml_introduction.ipynb Links
- InternalError: Dst tensor is not initialized. [[{{node IteratorGetNext/_2}}]] [Op:__inference_distributed_function_24557]
- In the tf1 example: I replace the weigtht and bias with tf.layers.dense, I found the accuracy decrease. why??? HOT 1
- The CNN example diagram shows 3 conv & pooling layers but the implementation only has 2
- AttributeError on placeholder HOT 1
- fig
- In K-Means Example, when i am running "from tensorflow.contrib.factorization import KMeans" line, i am getting an error "ModuleNotFoundError: No module named 'tensorflow.contrib'" HOT 2
- possible issue at: tensorflow_v2/notebooks/3_NeuralNetworks/autoencoder.ipynb HOT 6
- Add a development container HOT 3
- Ikvvh
- TPU Usage
- Activities HOT 3
- Offo HOT 3
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from tensorflow-examples.