dreamerv2
neptune-client
Our great sponsors
dreamerv2 | neptune-client | |
---|---|---|
4 | 24 | |
853 | 531 | |
- | 7.0% | |
0.0 | 9.6 | |
over 1 year ago | 7 days ago | |
Python | Python | |
MIT License | 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.
dreamerv2
-
Sources of Actor Gradients
In fact, they found that just reinforce gradients work in DM control now too: Dreamerv2 GitHub (they just needed to turn off gradients through the action path - which I guess was being passed back with straight-through estimation? I'm actually having a difficult time telling how the gradient is different on the action vs policy.log_prob(action)).
-
PyDreamer: model-based RL written in PyTorch + integrations with DM Lab and MineRL environments
This is my implementation of Hafner et al. DreamerV2 algorithm. I found the PlaNet/Dreamer/DreamerV2 paper series to be some of the coolest RL research in recent years, showing convincingly that MBRL (model-based RL) does work and is competitive with model-free algorithms. And we all know that AGI will be model-based, right? :)
-
Any current state or the art libraries for training agents to play atari games?
Last I checked, for running off a single node, the state of the art was Dreamerv2 https://github.com/danijar/dreamerv2
- Google AI, DeepMind And The University of Toronto Introduce DreamerV2, The First Reinforcement Learning (RL) Agent That Outperforms Humans on The Atari Benchmark
neptune-client
-
Show HN: A gallery of dev tool marketing examples
Hi I am Jakub. I run marketing at a dev tool startup https://neptune.ai/ and I share learnings on dev tool marketing on my blog https://www.developermarkepear.com/.
Whenever I'd start a new marketing project I found myself going over a list of 20+ companies I knew could have done something well to ācopy-pasteā their approach as a baseline (think Tailscale, DigitalOCean, Vercel, Algolia, CircleCi, Supabase, Posthog, Auth0).
So past year and a half, Iāve been screenshoting examples of how companies that are good at dev marketing do things like pricing, landing page design, ads, videos, blog conversion ideas. And for each example I added a note as to why I thought it was good.
Now, it is ~140 examples organized by tags so you can browse all or get stuff for a particular topic.
Hope it is helpful to some dev tool founders and marketers in here.
wdyt?
Also, I am always looking for new companies/marketing ideas to add to this, so if youād like to share good examples Iād really appreciate it.
-
How to structure/manage a machine learning experiment? (medical imaging)
There are a lot of tools out there for experiment tracking (eg neptune.ai), but I'm really not sure whether that sort of thing is over the top for what I need to do.
-
How to grow a developer blog to 3M annual visitors? with Jakub Czakon (Neptune.ai)
Welcome to another episode of The Developer-led Podcast, where we dive into the strategies modern companies use to build and grow their developer tools. In this exciting episode, we're joined by Jakub Czakon, the CMO at Neptune.ai, a startup that assists developers in efficiently managing their machine-learning model data. Jakub is renowned not only for his role at Neptune.ai but also for his developer marketing endeavors, including the influential newsletter Developer Markepear and a thriving developer marketing Slack community.
-
[D] Is there any all in one deep learning platform or software
tbh I have done a pretty good search on this topic, I couldn't find any. I thought maybe community could help me find one, if people like you (who works at neptune.ai) have the same opinion then it is what it is :). anyway thank you for the suggestions that you gave, probably gonna use that.
-
New Data Scientist, want to get into MLOps, where to start?
To get started with MLOps, you will need to have some foundational skills in Python, SQL, mathematics, and machine learning algorithms and libraries. You will also need to learn about databases, model deployment, continuous integration, continuous delivery, continuous monitoring, and other best practices of MLOps. You can find some useful resources for each of these topics in the following blogs on neptune.ai (disclosure: I work for Neptune):
-
Does a fully sentient (Or at least as sentient as you and me) AI with free will have a soul?
arxiv.org2. apro-software.com3. en.wikipedia.org4. neptune.ai
-
[D] The hype around Mojo lang
Other companies followed the same route to promote their paid product, e.g. plotly -> dash, Pytorch Lightning -> Lightning AI, run.ai, neptune.ai . It's actually a fair strategy, but some people may fear the conflict of interest. Especially, when the tools require some time investment, and it seems like a serious vendor lock-in. Investing some time to learn a tool is not such a big deal, but once you adapt a workflow of an entire team it can be tough to go back.
-
[P] New Open Source Framework and No-Code GUI for Fine-Tuning LLMs: H2O LLM Studio
track and compare your model performance visually. In addition, Neptune integration can be used.
-
[D] New features and current problems with ml infrastructure?
I am working on a startup, I was wondering what people think are some gaps in current machine learning infrastructure solutions like WandB, or Neptune.ai.
- All your ML model metadata in a single place
What are some alternatives?
dreamerv3 - Mastering Diverse Domains through World Models
MLflow - Open source platform for the machine learning lifecycle
dreamer - Dream to Control: Learning Behaviors by Latent Imagination
Serpent.AI - Game Agent Framework. Helping you create AIs / Bots that learn to play any game you own!
panda-gym - Set of robotic environments based on PyBullet physics engine and gymnasium.
Caffe - Caffe: a fast open framework for deep learning.
dm_control - Google DeepMind's software stack for physics-based simulation and Reinforcement Learning environments, using MuJoCo.
Pytorch - Tensors and Dynamic neural networks in Python with strong GPU acceleration
stable-baselines3-contrib - Contrib package for Stable-Baselines3 - Experimental reinforcement learning (RL) code
mxnet - Lightweight, Portable, Flexible Distributed/Mobile Deep Learning with Dynamic, Mutation-aware Dataflow Dep Scheduler; for Python, R, Julia, Scala, Go, Javascript and more
planet - Learning Latent Dynamics for Planning from Pixels
Porcupine Ā - On-device wake word detection powered by deep learning