COMMITS
/ torchao/quantization/quant_primitives.py August 13, 2025
A
Add float8 FakeQuantizeConfig and FakeQuantizer (#2735)
andrewor14 committed
A
Drop support for PyTorch 2.5 and before (#2720)
andrewor14 committed
August 4, 2025
V
fix float8 rowwise inference perf with torch.compile (#2672)
Vasiliy Kuznetsov committed
July 25, 2025
J
Misc fixes to prepare for adding Float8Tensor (#2603)
Jerry Zhang committed
July 23, 2025
S
add TORCH_VERSION_CHECK for _register_meta (#2575)
shiyang-weng committed
July 9, 2025
R
TorchAO new observers (#2508)
Rohan Joshi committed
July 1, 2025
A
Register choose_qparams_affine_float8 as custom op (#2461)
Angela Yi committed
June 25, 2025
S
June 18, 2025
A
[BE] Convert quant_primitives methods private (#2350)
Apurva Jain committed
June 10, 2025
A
[BE] Rename qparams for tinygemm (#2344)
Apurva Jain committed
June 5, 2025
Y
primitive scale fix
YIWENX14 committed
June 3, 2025
A
Fix QAT range learning, ensure scales get gradients (#2280)
andrewor14 committed
May 27, 2025
D
Fix Per Row scaling for inference (#2253)
Driss Guessous committed
May 21, 2025
A
Remove preserve_zero and zero_point_domain from choose_qparams_affine (#2149)
Apurva Jain committed
April 22, 2025
J
Update ruff version to 0.11.6 (#2103)
Jerry Zhang committed
April 15, 2025
J
Copy and adapt pt2e quantization code to torchao (#2048)
Jerry Zhang committed
April 10, 2025
M
INT4 XPU enabling (#1577)
Meng, Hengyu committed
April 8, 2025
J
Add gguf q4_k quantization (#2001)
Jerry Zhang committed
January 29, 2025
S
Consolidate `ZeroPointDomain.NONE` & `None` zero point domains (#1556)
sanchitintel committed
January 25, 2025
D
add separate quantization primitives for float8 (#1597)
Daniel Vega-Myhre committed
January 14, 2025
A
Fix torch.intx support in FakeQuantizeConfig (#1544)
andrewor14 committed
December 16, 2024
H
gemlite integration in torchao (#1034)
HDCharles committed
November 14, 2024
H
support W4A8 Marlin kernel (#1113)
HandH1998 committed
November 13, 2024
J
Update torchao api reference and add contributor guide (#1255)
Jerry Zhang committed
November 12, 2024
A
Ruff Lint fixes (torchao/quantization) (#1244)
Apurva Jain committed
November 4, 2024
T
BF16 support for Quant-LLM kernel (#1147)
Tobias van der Werff committed
October 30, 2024
S
Subclass API (#966)
Scott Roy committed
October 15, 2024
A
Add generic fake quantized linear for QAT (#1020)
andrewor14 committed
October 10, 2024
A
Move and rename GranularityType -> Granularity (#1038)
andrewor14 committed
September 12, 2024
J
Renaming fpx to floatx (#877)
Jerry Zhang committed
September 7, 2024
M
Update the way scale is calculated for affine Symmetric (#805)
Mengtao Yuan committed
September 6, 2024
J
Expose hqq through `uintx_weight_only` API (#786)
Jerry Zhang committed
D
[StaticQuant] add a linear observer class and test (#807)
Driss Guessous committed
September 5, 2024
A
Add static quant to float8 (#787)
Apurva Jain committed
D
[StaticQuant] Update how block_size is calculated with Observers (#815)
Driss Guessous committed
September 4, 2024
J
Add more information to quantized linear module and added some logs (#782)
Jerry Zhang committed
August 30, 2024
D
Add Float8 Weight Only and FP8 weight + dynamic activation (#740)
Driss Guessous committed
August 29, 2024
J
[reland] Refactor quant_llm to work with affine quantized tensor (#696) (#772)
Jerry Zhang committed
August 28, 2024
A
[Experimental] Float8 support in AQT (#671)
Apurva Jain committed
J
Revert "Refactor quant_llm to work with affine quantized tensor" (#767)
Jerry Zhang committed
J
Refactor quant_llm to work with affine quantized tensor (#696)
Jerry Zhang committed
August 23, 2024
J
Use `torch.uint1` to `torch.uint7` for Uintx tensor subclass (#672)
Jerry Zhang committed
S
Fix docstring args names (#735)
Sergii Dymchenko committed
August 15, 2024
M
Add HQQ support (#605)
mobicham committed
August 14, 2024
M
retry version guard fix (#679)
Mark Saroufim committed
J
Add AffineQuantizedObserver (#650)
Jerry Zhang committed
August 8, 2024
J
Add support for using AffineQuantizedTensor with `weights_only=True` (#630)
Jerry Zhang committed
August 3, 2024
J
Unpin nightly version (#593)
Jerry Zhang committed
July 17, 2024
A
Add cachemask variant for fake_quantize_affine (#500)
andrewor14 committed
July 11, 2024
A
Add general fake_quantize_affine op (#492)
andrewor14 committed