Automating Deep Learning
Tianqi reflects on his journey from building deep learning frameworks to realizing the bottlenecks in the process. He emphasizes the challenges of adapting to new model variants and the need for automation to streamline development. His experience at GTC sparked the idea of creating a more agnostic solution, leading to a two-year effort to build a comprehensive compilation system from the ground up.In this clip
From this podcast
Open Source Startup Podcast
E114: How OctoML Helps Developers Build with Llama 2 & Stable Diffusion
Related Questions