COMMITS
/ torchao/ops.py August 13, 2025
A
Drop support for PyTorch 2.5 and before (#2720)
andrewor14 committed
June 25, 2025
X
[CPU] Enable DA8W4 on CPU (#2128)
Xia Weiwen committed
June 12, 2025
J
Fix 2:4 sparsify meta registrations (#2366)
Jesse Cai committed
May 22, 2025
J
May 21, 2025
X
Re-land the PR of "Add INT8 SDPA path for CPU" (#2215)
Xuan Liao committed
May 12, 2025
J
2:4 activation sparsity packing kernels (#2012)
Jesse Cai committed
May 6, 2025
A
Fix linux cpu builds. Resolves nightly build for mac stops on 0422 (#2170)
Andrey Talman committed
May 2, 2025
D
[MX] Remove mxfp8 kernel and rely on cublas (#2130)
Driss Guessous committed
April 28, 2025
X
Re-land "Add INT8 SDPA path for CPU" (#2093)
Xuan Liao committed
April 22, 2025
J
[reland][ROCm] preshuffled weight mm (#2044)
Jeff Daily committed
J
Revert "Add INT8 SDPA path for CPU" (#2092)
Jerry Zhang committed
April 18, 2025
X
Add INT8 SDPA path for CPU (#1372)
Xuan Liao committed
April 9, 2025
J
Revert "[ROCm] preshuffled weight mm" (#2031)
Jerry Zhang committed
April 8, 2025
J
[ROCm] preshuffled weight mm (#1702)
Jeff Daily committed
April 2, 2025
D
Allow builds on less than sm75 raise runtime failure (#1999)
Driss Guessous committed
March 17, 2025
A
Add copyright header and pre-commit validator (#1900)
Apurva Jain committed
March 13, 2025
A
Add CUTLASS-based row-wise scaled sparse FP8 kernel (#1671)
Aleksandar Samardžić committed
March 5, 2025
V
enable compile with mxfp8 and mxfp4 cutlass gemm (#1838)
Vasiliy Kuznetsov committed
February 22, 2025
D
Allow for scales to be in new e8m0 dtype (#1742)
Driss Guessous committed
February 19, 2025
J
promote blocksparse from prototype, make it faster (#1734)
Jesse Cai committed
February 12, 2025
D
Add mx_fp8_bf16 kernel (#1637)
Driss Guessous committed
February 5, 2025
T
Add CUTLASS-based W4A4 (#1515)
Thien Tran committed
January 5, 2025
A
W4A8 based on CUTLASS (#880)
Aleksandar Samardžić committed
December 2, 2024
A
Lint fixes torchao files (#1366)
Apurva Jain committed
November 14, 2024
H
support W4A8 Marlin kernel (#1113)
HandH1998 committed
November 4, 2024
T
BF16 support for Quant-LLM kernel (#1147)
Tobias van der Werff committed
September 26, 2024
T
Move op definition for custom kernel from C++ to Python (#949)
Thien Tran committed
September 12, 2024
J
Renaming fpx to floatx (#877)
Jerry Zhang committed
August 23, 2024
D
Add sparse marlin 2:4 gemm op (#733)
Diogo Venâncio committed
S
Fix docstring args names (#735)
Sergii Dymchenko committed
August 14, 2024
M
retry version guard fix (#679)
Mark Saroufim committed
July 4, 2024
J
[FEAT] Add custom CUDA `tinygemm` unpacker (#415)
jeromeku committed
June 25, 2024
T
Add FP5 E2M2 support from upstream (#399)
Thien Tran committed
June 13, 2024
T
Add FP6-LLM doc and move FP6-LLM to prototype (#358)
Thien Tran committed
June 10, 2024
T
FP6-LLM clean up (again) (#339)
Thien Tran committed
June 9, 2024
M
Revert "Clean up FP6-LLM" (#338)
Mark Saroufim committed
T
Clean up FP6-LLM (#304)
Thien Tran committed
June 7, 2024
J
Move some util functions from quantization.utils to torchao.utils (#337)
Jerry Zhang committed
May 25, 2024
T
Improve primitives for FP6 quant (#248)
Thien Tran committed
May 24, 2024
D
Fix Readme and remove unused kernel (#270)
Driss Guessous committed
M
Remove test_ops.py warning spew (#267)
Mark Saroufim committed
May 14, 2024
T
Add FP16Act-FP6Weight Linear (#223)
Thien Tran committed
April 26, 2024
M
Custom CUDA extensions (#135)
Mark Saroufim committed