|
About the compiler category
|
|
0
|
452
|
January 22, 2021
|
|
TorchInductor Update 4: CPU backend started to show promising performance boost
|
|
1
|
204
|
November 25, 2022
|
|
Dynamo/FX: patching a function to add more outputs not working
|
|
2
|
78
|
November 21, 2022
|
|
State of symbolic shapes branch
|
|
18
|
1278
|
November 21, 2022
|
|
TorchDynamo Update 9: Making DDP Work with TorchDynamo
|
|
1
|
354
|
November 18, 2022
|
|
TorchDynamo: An Experiment in Dynamic Python Bytecode Transformation
|
|
6
|
4500
|
November 15, 2022
|
|
TorchDynamo Update 3: GPU Inference Edition
|
|
10
|
2328
|
November 15, 2022
|
|
Where we are headed and why it looks a lot like Julia (but not exactly like Julia)
|
|
8
|
27669
|
November 14, 2022
|
|
The nuances of PyTorch Graph Capture
|
|
5
|
2883
|
November 7, 2022
|
|
Min-cut optimal(*) recomputation (i.e. activation checkpointing) with AOTAutograd
|
|
6
|
2928
|
January 20, 2022
|
|
Skipping Dispatcher with LazyTensor
|
|
10
|
555
|
October 19, 2022
|
|
TorchInductor Update 3: E2E model training with TorchDynamo + Inductor gets 1.67x/2.1x speedup
|
|
3
|
632
|
October 11, 2022
|
|
TorchInductor: a PyTorch-native Compiler with Define-by-Run IR and Symbolic Shapes
|
|
17
|
3413
|
October 10, 2022
|
|
Reducing Framework Overhead with Static Runtime
|
|
4
|
912
|
October 4, 2022
|
|
Tracing with Primitives: Update 0
|
|
17
|
2376
|
September 26, 2022
|
|
NNC walkthrough: how PyTorch ops get fused
|
|
10
|
2666
|
November 3, 2021
|
|
TorchDynamo Update 8: TorchDynamo passed correctness check on 7k+ github models
|
|
7
|
820
|
July 1, 2022
|
|
Tracing with Primitives: Update 2
|
|
3
|
2394
|
June 25, 2022
|
|
TorchDynamo Update 7: Inference with FX2TRT
|
|
1
|
937
|
June 16, 2022
|
|
Python Operator Authoring w/ NNC
|
|
5
|
1253
|
June 7, 2022
|
|
Where do the 2000+ PyTorch operators come from?: More than you wanted to know
|
|
12
|
3920
|
May 12, 2022
|
|
Tracing with Primitives: Update 1, nvFuser and its Primitives
|
|
0
|
1188
|
April 25, 2022
|
|
TorchDynamo Update 5: Improved Capture & Bigger Graphs
|
|
4
|
1111
|
April 14, 2022
|
|
TorchDynamo Update 6: Training support with AOTAutograd
|
|
0
|
1288
|
March 29, 2022
|
|
Universal binaries for libtorch Mac?
|
|
10
|
1723
|
March 14, 2022
|
|
TorchDynamo Update 4: LazyTensor & nvFuser Experiments
|
|
2
|
1598
|
February 25, 2022
|
|
Prim::autocast_promote operation?
|
|
4
|
541
|
December 31, 2021
|
|
TorchDynamo Update: 1.48x geomean speedup on TorchBench CPU Inference
|
|
0
|
1924
|
November 12, 2021
|
|
Next Steps for PyTorch Compilers
|
|
9
|
4225
|
October 21, 2021
|
|
Loop_tool's lazy frontend - Experimenting with symbolic laziness
|
|
0
|
317
|
October 11, 2021
|