Treat the initial state(s) of TensorFlow Keras recurrent neural network (RNN) layers as a parameter or parameters to be learned during training, as recommended in, e.g., [1].
Ordinary RNNs use an all-zero initial state by default. Why not let the neural network learn a smarter initial state?
The trainable-initial-state-rnn package provides a class TrainableInitialStateRNN that can wrap any tf.keras RNN (or bidirectional RNN) and manage new initial state variables in addition to the RNN’s weights.
Typical usage looks as follows.
import tensorflow as tf
from trainable_initial_state_rnn import TrainableInitialStateRNN
base_rnn = tf.keras.layers.LSTM(256)
rnn = TrainableInitialStateRNN(base_rnn) # Treats initial state as a variable!
model = tf.keras.Model(...) # Use rnn like any other tf.keras layer in your model
model.compile(...)
history = model.fit(...)Documentation is available at Read the Docs.
Installation
The trainable_initial_state_rnn package can be installed using the pip utility directly from the package’s GitHub page:
pip install git+https://github.com/artemmavrin/trainable-initial-state-rnn.gitAlternatively, install a recent release from the Python Package Index (PyPI):
pip install trainable-initial-state-rnnNote. To install the project for development (e.g., to make changes to the
source code), clone the project repository from GitHub and run make dev:
git clone https://github.com/artemmavrin/trainable-initial-state-rnn.git
cd trainable-initial-state-rnn
# Optional but recommended: create and activate a new Python virtual environment
make devThis will additionally install the requirements needed to run tests, check code coverage, and produce documentation.
References
1 maintainer
Release files for trainable-initial-state-rnn 0.0.3
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| trainable-initial-state-rnn-0.0.3.tar.gz | 6.9 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| trainable_initial_state_rnn-0.0.3-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 16.9 kB
Release files / trainable-initial-state-rnn-0.0.3.tar.gz
| Download URL | trainable-initial-state-rnn-0.0.3.tar.gz |
|---|---|
| Size | 6.9 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
eed8dd6c1ee98739910580ef1285d298e1f92109cd8688d3fb99e6fbcef0c735
|
|
BLAKE2b-256 checksum How to use checksums |
acd5041e82e263e87273d00b56ada6d83d4c3bf005420294f966e80c83865e18
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/3.1.1 pkginfo/1.5.0.1 requests/2.23.0 setuptools/47.1.1 requests-toolbelt/0.9.1 tqdm/4.46.1 CPython/3.8.3
|
Release files / trainable_initial_state_rnn-0.0.3-py3-none-any.whl
| Download URL | trainable_initial_state_rnn-0.0.3-py3-none-any.whl |
|---|---|
| Size | 10.0 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
be3d92098f7cf5b05ee2088d0897fe35f591217407393a842d5b84a07ef841b2
|
|
BLAKE2b-256 checksum How to use checksums |
b6b79dcbab790394844366ce572d3d74fa2eaf37085a107e9890b76b3b232f81
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/3.1.1 pkginfo/1.5.0.1 requests/2.23.0 setuptools/47.1.1 requests-toolbelt/0.9.1 tqdm/4.46.1 CPython/3.8.3
|
