Practical Deep Learning with Rachel Thomas - TWiML Talk #138

This article summarizes a podcast episode featuring Rachel Thomas, founder of Fast AI. The discussion centers around Fast AI's educational courses, particularly "Practical Deep Learning for Coders." The conversation covers the philosophy behind the courses, designed to make deep learning accessible without requiring extensive mathematical prerequisites. Key topics include Fast AI's shift from TensorFlow to PyTorch, the rationale behind this decision, and the lessons learned. The article also highlights the Fast AI deep learning library and its role in achieving significant improvements in training time and cost on an industry benchmark. The focus is on practical applications and accessibility of deep learning.

• •Fast AI offers courses to make deep learning accessible to those without extensive math backgrounds.
• •Fast AI switched from TensorFlow to PyTorch for its courses.
• •The Fast AI deep learning library helped achieve significant improvements in training time and cost.