PyTorch Developer Podcast
By Edward Yang, Team PyTorch
Listen to a podcast, please open Podcast Republic app. Available on Google Play Store and Apple App Store.
Image by Edward Yang, Team PyTorch
Category: Technology
Open in Apple Podcasts
Open RSS feed
Open Website
Rate for this podcast
Subscribers: 22
Reviews: 0
Episodes: 83
Description
The PyTorch Developer Podcast is a place for the PyTorch dev team to do bite sized (10-20 min) topics about all sorts of internal development topics in PyTorch.
| Episode | Date |
|---|---|
|
Compiler collectives
|
Aug 04, 2024 |
|
TORCH_TRACE and tlparse
|
Apr 29, 2024 |
|
Higher order operators
|
Apr 21, 2024 |
|
Inductor - Post-grad FX passes
|
Apr 12, 2024 |
|
CUDA graph trees
|
Mar 24, 2024 |
|
Min-cut partitioner
|
Mar 17, 2024 |
|
AOTInductor
|
Mar 02, 2024 |
|
Tensor subclasses and PT2
|
Feb 24, 2024 |
|
Compiled autograd
|
Feb 19, 2024 |
|
PT2 extension points
|
Feb 05, 2024 |
|
Inductor - Define-by-run IR
|
Jan 24, 2024 |
|
Unsigned integers
|
Jan 17, 2024 |
|
Inductor - IR
|
Jan 16, 2024 |
|
Dynamo - VariableTracker
|
Jan 12, 2024 |
|
Unbacked SymInts
|
Feb 21, 2023 |
|
Zero-one specialization
|
Feb 20, 2023 |
|
torchdynamo
|
Dec 06, 2022 |
|
PyTorch 2.0
|
Dec 04, 2022 |
|
History of functorch
|
Nov 07, 2022 |
|
Learning rate schedulers
|
Jun 13, 2022 |
|
Weak references
|
Jun 06, 2022 |
|
Strides
|
May 30, 2022 |
|
AOTAutograd
|
May 09, 2022 |
|
Dispatcher questions with Sherlock
|
May 02, 2022 |
|
New CI
|
Apr 25, 2022 |
|
Python exceptions
|
Apr 17, 2022 |
|
Torch vs ATen APIs
|
Apr 11, 2022 |
|
All about NVIDIA GPUs
|
Sep 24, 2021 |
|
Tensor subclasses and Liskov substitution principle
|
Sep 16, 2021 |
|
Half precision
|
Sep 10, 2021 |
|
DataLoader with multiple workers leaks memory
|
Sep 01, 2021 |
|
Batching
|
Aug 18, 2021 |
|
Multiple dispatch in __torch_function__
|
Aug 10, 2021 |
|
Multithreading
|
Aug 03, 2021 |
|
Asynchronous versus synchronous execution
|
Jul 27, 2021 |
|
gradcheck
|
Jul 23, 2021 |
|
torch.use_deterministic_algorithms
|
Jul 21, 2021 |
|
Reference counting
|
Jul 20, 2021 |
|
Memory layout
|
Jul 13, 2021 |
|
pytorch-probot
|
Jul 12, 2021 |
|
API design via lexical and dynamic scoping
|
Jul 09, 2021 |
|
Intro to distributed
|
Jul 08, 2021 |
|
Double backwards
|
Jul 07, 2021 |
|
Functional modules
|
Jul 06, 2021 |
|
CUDA graphs
|
Jun 28, 2021 |
|
Default arguments
|
Jun 25, 2021 |
|
Anatomy of a domain library
|
Jun 24, 2021 |
|
TensorAccessor
|
Jun 23, 2021 |
|
Random number generators
|
Jun 22, 2021 |
|
vmap
|
Jun 21, 2021 |
|
Expect tests
|
Jun 18, 2021 |
|
XLA
|
Jun 17, 2021 |
|
TH
|
Jun 16, 2021 |
|
TorchScript
|
Jun 15, 2021 |
|
CMake
|
Jun 14, 2021 |
|
torchdeploy
|
Jun 11, 2021 |
|
C++ frontend
|
Jun 10, 2021 |
|
PyObject preservation
|
Jun 09, 2021 |
|
Mobile selective build
|
Jun 08, 2021 |
|
torch.nn
|
Jun 07, 2021 |
|
Code generation
|
Jun 04, 2021 |
|
Why is autograd so complicated
|
Jun 03, 2021 |
|
__torch_function__
|
Jun 02, 2021 |
|
TensorIterator
|
Jun 01, 2021 |
|
native_functions.yaml
|
May 28, 2021 |
|
Serialization
|
May 27, 2021 |
|
Continuous integration
|
May 26, 2021 |
|
Stacked diffs and ghstack
|
May 25, 2021 |
|
Shared memory
|
May 24, 2021 |
|
Automatic mixed precision
|
May 21, 2021 |
|
Conjugate views
|
May 20, 2021 |
|
History and constraints of Tensor
|
May 19, 2021 |
|
How new operators are authored
|
May 18, 2021 |
|
The life and death of Variable
|
May 17, 2021 |
|
Backend extensibility
|
May 14, 2021 |
|
The road to structured kernels
|
May 13, 2021 |
|
Functionalization
|
May 12, 2021 |
|
Just enough CUDA to be dangerous
|
May 11, 2021 |
|
Inference mode
|
May 10, 2021 |
|
Vectorization
|
May 07, 2021 |
|
Dynamic library structure
|
May 06, 2021 |
|
History and constraints of the dispatcher
|
May 05, 2021 |
|
Binding C++ objects to Python
|
May 04, 2021 |