- 13
- 24 051
Daniel Eid
Приєднався 3 лис 2012
Check out some of my work at danieleidworks.com
github.com/danieleid317
Looking for paid work so hmu if im doing some things you might need for your business.
github.com/danieleid317
Looking for paid work so hmu if im doing some things you might need for your business.
Calculating how frequently a prime is the smallest factor of all integers . Very prime integers .
We cover how to determine the likelihood that a prime is the smallest factor or smallest divisor of any integer . We then extend that to determine the likelihood that specific integers are divisible by smaller primes . A sample case to find very prime integers , is used to show its applicability .
github.com/danieleid317/prime
github.com/danieleid317/prime
Переглядів: 76
Відео
Scaling data-driven robotics with reward sketching in pytorch
Переглядів 1492 роки тому
arxiv.org/pdf/1909.12200.pdf danieleidworks.com/ github.com/danieleid317
Deep Reinforcement Learning From Human Preferences in tensorflow
Переглядів 1,7 тис.2 роки тому
A simple implementation of using human preferences to train a reward function on user preference given agent data. arxiv.org/pdf/1706.03741.pdf www.danieleidworks.com/ github.com/danieleid317
Django models , forms , database , and working with user data
Переглядів 402 роки тому
You learn how to create data models, migrate a database, create forms form those models, and use those forms to accept user data and add it to a database as well as querying a database. This is all done to build a sample application.
Serve a simple webpage using django with python
Переглядів 392 роки тому
2022 02 17 13 07 18 Start a basic django project and learn how to handle requests and return responses.
Pytorch Continuous A2C RNN agent, bonus sample code, and 'Deep Mimic' multi skill decoder tinkering
Переглядів 7132 роки тому
2022 02 08 13 09 51 I go over the implementation of using a simple RNN inside of a training loop for an RL environment. I also discuss uses of GRU to improve the network and personal studies/tests to modify Deep Mimic for recurrent networks.
pybullet for custom python RL environments
Переглядів 18 тис.2 роки тому
2022 02 02 12 54 13 Here we learn the basics of pybullet and setting it up in a custom python environment. danieleidworks.com/ github.com/danieleid317
Custom python env, discrete REINFORCE in tensorflow
Переглядів 1892 роки тому
2022 01 31 14 43 52 Baby steps in the application of RL on custom python environments using tensorflow.
Discrete REINFORCE in tensorflow and openai gym
Переглядів 1692 роки тому
2022 01 30 15 41 53 We train a tensorflow model to maximize future expected sum of rewards for the CartPole environment in openai gym.
pytorch.distributions and tensorflow_probability
Переглядів 5632 роки тому
2022 01 30 14 29 48 Hello , this is a quick guide on how to take network output and convert it into a policy distribution used for sampling, log likelihood , and entropy measure. Also i don't edit my videos and do them all in one shot, so sorry for any hiccups that happen along the way. github.com/danieleid317 www.danieleidworks.com/
Build a custom RL environment in python
Переглядів 1,4 тис.2 роки тому
2022 01 28 11 46 08 Tensorflow , openai gym , and other packages are available to help you build custom environments for your reinforcement learning agents. I show you how to build a simple class with only a few class methods in order to give you granular control over your training in python.
Training a neural network in tensorflow
Переглядів 512 роки тому
2022 01 27 11 03 02 Being fast and loose with the first few tutorials. Runnable code segment
Intentions and expectations for videos
Переглядів 312 роки тому
2022 01 26 22 26 37 Introducing that this will be a runnable code channel.
Hell yeah dude
That's how I feel . Isn't this incredible ?
So I’m not all the way through it yet, but it’s really interesting so far. You should look into applied cymatics if you like emergent patterns like this.
@ yeah I’ll look into it. Always interested in different math and patterns (:
@@danieleid8528right on man. I’m with a company that intends to use cymatic principles for manufacturing. We’re about to build a facility in Arizona. So let me know if it piques your interest.
so nice video
is the link to your code available anywhere?
In the video. You should be transcribing following along locally or in a colab notebook.
Thanks for the video.
Excellent video, thank you!
where is the GitHub repository for the code, Sir
Hey, I wanted to ask a question / point something out. In the prefferenc update section you take the next state for the rewarder to produce a reward on "reward1 = rewarder(states[transition_ids[0] + 1])", But I imagine that would mean the rewarder model is only learning from the states and not the state action pair as seen in the paper. Also I was curious why you chose the next state by adding the + 1 because your states are already storing the next states based on where you are appending the states in the step function. So in a way by adding +1 you are going to the next next state. Anyways these were my confusions, get back to me if possible
this is great dude, very helpful. Im getting into rl and implementing it in my field (architecture) but i dont have a computer science or math backgound, so when i read this paper for the first time, i was quite confused. However with the help of your explanantion in this video, it all makes much more sense now. Thanks
The video is all I needed to get started, amazing! Thanks for the effort
This is a great tutorial man, thank you. Also is there a group or opensource project community where I can learn more about this? Actually I am a reinforcement learning enthusiast and want to apply it here.
What a poorly made video at the end, after the walkthrough was done):
Thank you
Great tutorial man! Keep it up
I can't find your code in the video can someone send the link to it.
Hey did you find the code?
@@japneetsingh5015 Did anyone find the code?
Thank you, how do i add custom .obj files into the environment? I am working on robot that has to pick a custom object.
it's really helpful thank you so much!!
this is very helpful, thank you for this
Hey Daniel, thanks for the video again, would you like to make a futher video for how to apply RL or IRL using custom env (pybullet)? like combining the procedures of custom env and RL
Hi, i am currently working on that very topic, but do not expect to have it released for at least a couple months because i barely begun, have very little time to work on it, and it takes quite a bit of time. Please let me know what papers and algorithms you are referring to as far as IRL so i can review them in addition to the resources that i am using.
@@danieleid8528 Hi Daniel, really gald to recevie your reply. I am also working on that topic now, and im trying to apply IRL for robotic application, like grasping or peg-in-hole or some other touch-rich manipulation tasks. I found AIRL was quite interesting cos it can get expert's policy just using actions and observation. the paper im focusing on is 'learning robust rewards with adversarial inverse reinforcement learning', here is the link : arxiv.org/abs/1710.11248 best regards
Thanks for the tutorial
This was useful! Thanks
Thank you very much :)
What about customizing render( ) method? I cannot find anything on this online or anywhere for that matter. It would be a good tutorial if you can show how to visualize your own data with say a Matplotlib plot using render within jupyter notebooks.
Relay thanks man it's a great tutorial
Best tutorial about this topic so far!
Bro code plz
Great tutorial!