Comments (4)
Thanks for the quick reply @hardmaru, much appreciated.
In regards to point 1, I appear to have misinterpreted the paper. I did notice that the algorithm doesn't appear to have any scaling for the standard deviation term either, but your implementation seems to scale by the batch size?
delta_sigma = (np.dot(rS, S)) / (2 * self.batch_size * stdev_reward)
Anyway, ultimately, I don't think that it's very important. Was there another paper that had the additional tricks that you implemented (annealing learning rates etc.) or you just pulled them from the standard machine learning approaches?
You're quite right on point 2, I'm not sure how I interpreted that as a uniform sample.
cheers
from estool.
Hi @cammckenzie
Thanks for the message and interest.
For your first question regarding normalized by batch size, in the PEPG paper, in the right side of "Algorithm 1" on page 7 that I tried to base the implementation on, there is no scaling for population size. I guess the difference can be adjusted by adjusting the learning rate.
For the second point, I am not sampling from a uniform distribution, but sampling from a normal distribution using np.random.randn
as you mentioned.
self.epsilon = np.random.randn(self.batch_size, self.num_params) * self.sigma.reshape(1, self.num_params)
Thanks
from estool.
You're right, I incorporated scaling by batch (or population) size so that the same learn rate parameter can be used for various different population sizes (I think the paper should have done that too, just an obvious thing to do). The annealing and other tricks are standard tricks in deep learning, I recommend going through OpenAI's ES paper as well - they used Adam rather than vanilla SGD for instance.
from estool.
Thanks again @hardmaru, looking forward to your next blog whatever that may be.
from estool.
Related Issues (20)
- bipedhard doesn't converge HOT 14
- Do the parameters need to be in [a,b] HOT 2
- [Weight decay] Should weight decay of model parameters instead ? HOT 3
- Can't run pre-trained biped models HOT 2
- No colour HOT 1
- Bug in train.py HOT 1
- The pre-trained Kuka model did not work HOT 2
- raise NotImplementedError HOT 4
- class OpenES, function ask: np.array mu come from update and compute step Adam optimizer with shape(popsize, numparams) and it was error when try to reshape mu.reshape(1, self.num_params) HOT 1
- in PEPG module change_mu = np.dot(rT, epsilon) where left side have dimention pop_size and right have dimention batch_size(half of pop_size) HOT 1
- Natural gradients for deep layers HOT 1
- PEPG and NES HOT 2
- Question for CMA-ES test part
- Bad score when training "KukaBulletEnv-v0"
- matplotlib isn't in requirements.txt HOT 1
- What is the best way to step through the code ?
- The meaning of ”rms_stdev“ function?
- Class OpenES, function ask. self.mu never gets updated. HOT 1
- bug in train.py HOT 16
- How to train "KukaBulletEnv-v0" ? HOT 6
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 estool.