vizier
keras-tuner
vizier | keras-tuner | |
---|---|---|
5 | 4 | |
1,173 | 2,825 | |
0.7% | 0.5% | |
9.3 | 7.8 | |
1 day ago | about 1 month ago | |
Python | Python | |
Apache License 2.0 | Apache License 2.0 |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
vizier
- [N] Google Open Sources Vizier, Hyperparameter + Blackbox Optimization Service at Scale
- Is there any premade evolutionary algorithm selecting optimal NN architectures in TensorFlow ?
- Google just open sourced its Vizier optimisation suite
- Python-based research interface for blackbox and hyperparameter optimization
keras-tuner
- Is there any premade evolutionary algorithm selecting optimal NN architectures in TensorFlow ?
-
What has priority in the performance?
If you are using tensorflow, you can do all of that with the very elegant (in my opinion) https://keras.io/keras_tuner/
-
Space Science with Python - Asteroids meet Deep Learning #10
today I'd like to show you how to optimize a Conv1D network using Keras-Tuner. It enables one to automatically test some pre-defined networks; or it applies Bayesian or Hyperband optimization to find the best model!
-
How to know how many layers (LSTM and dense) should I create and how to know the right parameters? (beginner)
Tbh, just try and error. There is no right or wrong. You can use the Hyperparameter Tuner from keras to define some architectures with varying number of layers and units as well as some other hyperparameters.
What are some alternatives?
mango - Parallel Hyperparameter Tuning in Python
Ray - Ray is a unified framework for scaling AI and Python applications. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.
SpaceDrones - A simple learning environment with space drones for evolution-inspired optimization.
deephyper - DeepHyper: Scalable Asynchronous Neural Architecture and Hyperparameter Search for Deep Neural Networks
Gradient-Free-Optimizers - Simple and reliable optimization with local, global, population-based and sequential techniques in numerical discrete search spaces.
nni - An open source AutoML toolkit for automate machine learning lifecycle, including feature engineering, neural architecture search, model compression and hyper-parameter tuning.
tune - An abstraction layer for parameter tuning
d2l-en - Interactive deep learning book with multi-framework code, math, and discussions. Adopted at 500 universities from 70 countries including Stanford, MIT, Harvard, and Cambridge.
mlr3hyperband - Successive Halving and Hyperband in the mlr3 ecosystem
baybe - Bayesian Optimization and Design of Experiments
DIgging - Decision Intelligence for digging best parameters in target environment.
pyswarms - A research toolkit for particle swarm optimization in Python