Software Engineering Radio - the podcast for professional software developers

Episode 549: William Falcon Optimizing Deep Learning Models

11 snips
Feb 3, 2023
Ask episode
Chapters
Transcript
Episode notes
1
Introduction
00:00 • 2min
2
The Challenges of Deep Learning
01:37 • 3min
3
Do You Think You Need Deep Learning?
04:27 • 2min
4
Optimizing Deep Learning Models
06:45 • 2min
5
Is Training Versus Inferencing a Thing?
09:15 • 3min
6
Is There a Transition Point for Distributed Training?
12:06 • 2min
7
MLOps Tasks - Can You Do Two Iterations a Day?
14:26 • 3min
8
Is There a Barrier to Entry in the Deep Learning Industry?
17:44 • 3min
9
How to Train GPUs in the Cloud?
20:56 • 3min
10
Getting Rid of Boilerplate in PyTorch
23:58 • 4min
11
Is There Any Pitfall in Deep Learning?
28:15 • 4min
12
Using Lightning to Train Machine Learning Models
31:55 • 2min
13
PyTorch - What's the Difference Between a Torch and a Module?
34:12 • 2min
14
How to Train, Validate, Test and Predict
36:38 • 2min
15
How to Use Lightning to Improve Your Model Performance
38:53 • 3min
16
Lightning and Data Versioning
41:52 • 2min
17
Debugging in Machine Learning?
43:54 • 2min
18
Debugging Distributed Training Bugs?
45:29 • 2min
19
Decoupling Data From the Model From the Hardware
47:24 • 2min
20
Is TensorFlow the Right Approach?
48:55 • 2min
21
Open-Source and the Needs of Academic Researchers Compared to Commercial Users
51:13 • 2min
22
Getting Started With PyTorch and Lightning
53:34 • 2min
23
Lightning - The Last Deep Learning Framework You'll Ever Need?
55:45 • 2min
24
Developing an Operating System for Artificial Intelligence
57:23 • 2min
25
The Next Big Thing Is AI, Right?
58:54 • 2min