Wild Technology

Fast.ai built more efficient Machine learning code

0

The future:-

The world is heading towards a technological era, an era where everything will be computerised no man force will be required for any kind of work. In the future robots will be responsible for lowering the labour work done by humans. To build robots to enter into the field of robotics one must know its basic need which is Artificial Intelligence. Machine learning is a vast topic, every Tech Savvy wants to learn and dig deeper and deeper into this. Whenever we think about Machine learning a name pops out of our mind “GOOGLE”. Google is the one who is digging more and more into Machine learning and is coming with high-end inventions, innovation. When it comes to machine learning then Google already has developed many high-end devices for eg. Google pods.

Introduction to the team:- 

Fast.ai is a group of people they run a small organization and have build Machine learning courses online. They are the one who has built a code which outperforms Google’s Machine learning code. Fast.ai is a team of students who have a keen interest in Machine learning. A group of students has built an algorithm that beats Google’s code as measured using a benchmark called DAWNBench. Benchmark uses a common image classification task to track the speed of a Deep learning algorithm per dollar of computing power. Google was till now the best in building custom chips specially designed for Machine learning. Fast.ai has surpassed it and are able to produce even faster on roughly equivalent hardware. Jeremy Howard & Rachel Thomas are the founder and co-founder of Fast.ai they made Fast.ai just to make artificial intelligence more accessible and less exclusive for common people.

Comparison between Fast.ai & Google:-

Howard and its team were able to compete with Google by doing a lot of simple things. Matei Zaharia, a professor at Stanford University and creator of the DAWNBench has said that Fast.ai work is impressive. The Fast.ai algorithm trained on the ImageNet database. The total computing cost for this was just $40 and it took only 18 minutes by using 16 Amazon web services. Founder claims that it is 40% better than Google’s effort. Although he admits comparison is different because both of us are using different hardware.

Air Pollution : How To Fight It With Nirvana Being idMask2

Previous article

Google launched a new Q&A app ‘Cameos’ focused on celebs

Next article

Comments

Leave a Reply

Login/Sign up