Tensorop
- 0 Collaborators
Tensorop is a Deep Learning library built over Pytorch to accelerate research in the simplest manner possible without compromising on flexibility. It provides support for vision and NLP. Currently in pre alpha,expect some bugs. The main idea is to provide the fundamental blocks and unified experience to the user which can then be inherited (built upon) for various purposes. ...learn more
Project status: Under Development
Groups
Student Developers for AI
Intel Technologies
Intel Opt ML/DL Framework
Overview / Usage
Tensorop is a Deep Learning library built over Pytorch to accelerate research in the simplest manner possible without compromising on flexibility. It will provide extensive support for vision, NLP and Reinforcement learning in a unified framework encompassing the best algorithms (state of the art) so that it makes it easy for researchers to implement various DL models/algorithms.
Documentations can be found here (https://github.com/prajjwal1/tensorop/tree/master/docs).
Methodology / Approach
Tensorop will inherit from Pytorch. It will have all fundamental classes for common tasks in vision/NLP/RL and then specialized algorithms will be designed to leverage those fundamentals.
It will include best practices as it's defaults + ongoing proven research algorithms for all kinds of tasks in CV/NLP/GANs/RL.
Technologies Used
Pytorch