In this post, I would like to share with you our development progress for the last month. Let’s check what features we have added to the framework in such a short time.
For real breakthroughs in deep learning, we need a strong foundation. In this blog post, I would like to introduce Catalyst framework, developed with focus on reproducibility, fast experimentation and code/idea reusing. We’ll also provide a tutorial on MNIST classification problem as an example.
In this episode, Sanyam Bhutani interviews a researcher, practitioner and open source contributor: Sergey Kolesnikov, creator of catalyst, which is a deep learning and reinforcement learning framework based on Pytorch.
We present Catalyst.RL, an open-source PyTorch framework for reproducible and sample efficient reinforcement learning (RL) research. Main features of Catalyst.RL include large-scale asynchronous distributed training, efficient implementations of …
Every year the topic of reinforcement learning (RL) is getting hotter and more hype. And every year, DeepMind and OpenAI add fuel to the fire with a new superhuman performance bot. Is there something really worthwhile behind this? And what are the latest trends in the entire RL variety? Let's find out!