issues
search
pytorch
/
ao
Native PyTorch library for quantization and sparsity
https://pytorch.org/ao
BSD 3-Clause "New" or "Revised" License
294
stars
41
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Add nightly build workflow
#250
huydhn
closed
1 week ago
8
Nightly Build Fix
#249
msaroufim
closed
1 week ago
2
Improve primitives for FP6 quant
#248
gau-nernst
closed
6 days ago
2
Remove args and kwargs from `AffineQuantizedTensor`
#247
jerryzh168
closed
2 weeks ago
2
Revert "Enable dispatch to tinygemm int4 and int8 kernels for quantized tensor"
#246
cpuhrsch
closed
2 weeks ago
3
Include installing requirements for building from source
#245
janeyx99
closed
2 weeks ago
1
Fix CI after quantize op change in PyTorch core
#244
andrewor14
closed
2 weeks ago
1
Remove input_quant_func from AffineQuantizedTensor subclass
#243
jerryzh168
closed
2 weeks ago
2
Fix CI after do_bench refactor in pytorch inductor
#242
jerryzh168
closed
2 weeks ago
1
python setup.py in build scripts
#241
msaroufim
closed
2 weeks ago
1
Update and rename nightly-build.yml to build.yml
#240
msaroufim
closed
2 weeks ago
1
[8da4w] Add support to skip quantizing some layers
#239
andrewor14
closed
2 weeks ago
4
In tutorials/quantize_vit, extract common methods to util.py
#238
lancerts
closed
1 week ago
5
[Question] MBU in automated CI?
#237
cadedaniel
opened
3 weeks ago
2
Small change to README.md
#236
svekars
closed
2 weeks ago
1
Skip depending on torch testing package
#235
jerryzh168
closed
3 weeks ago
2
Bump version to 0.2.0
#234
cpuhrsch
closed
3 weeks ago
1
Dedup _choose_qparams_per_token_asymmetric
#233
andrewor14
closed
2 weeks ago
3
Revert pyproject.toml changes
#232
msaroufim
closed
3 weeks ago
1
Building torchao from source installs unnecessary torch and nvidia packages every time
#231
gau-nernst
closed
3 weeks ago
1
Enable dispatch to tinygemm int4 and int8 kernels for quantized tensor
#230
jerryzh168
closed
2 weeks ago
4
Match torch.fake_quantize numerics in 8da4w QAT
#229
andrewor14
closed
2 weeks ago
5
FloatQuantization subclass
#228
msaroufim
opened
3 weeks ago
1
Add more options in choose_qparams_affine for tinygemm op
#227
jerryzh168
closed
3 weeks ago
2
Fix an error in subclass impl
#226
jerryzh168
closed
3 weeks ago
1
add expecttest to requirements.txt
#225
msaroufim
closed
3 weeks ago
1
Install dev-requirements.txt in doc build
#224
svekars
closed
3 weeks ago
4
Add FP16Act-FP6Weight Linear
#223
gau-nernst
closed
2 weeks ago
9
[BUG] No module named 'expecttest' when import `torchao`
#222
gau-nernst
closed
3 weeks ago
0
Doc build failing on main
#221
msaroufim
closed
3 weeks ago
0
Some follow up fixes for quant primitives
#220
jerryzh168
closed
3 weeks ago
1
[BE] Fix deprecation warning on torch.library
#219
msaroufim
closed
3 weeks ago
0
Don't import _C in fbcode
#218
jerryzh168
closed
3 weeks ago
0
Fix the URLs of web pages
#217
Jokeren
closed
3 weeks ago
0
Fused DoRA kernels
#216
jeromeku
closed
3 weeks ago
0
Update Regression Test to Unpinned Nightly
#215
msaroufim
closed
3 weeks ago
0
Unified AffineQuantizedTensor subclass
#214
jerryzh168
closed
3 weeks ago
0
Add test for torch.export.export
#213
jerryzh168
closed
4 weeks ago
0
Re-enable AOTI tests
#212
jerryzh168
closed
4 hours ago
4
Copy weights and preserve device for 8da4w QAT linear
#211
andrewor14
closed
3 weeks ago
2
Reduce memory usage for symmetric choose_qparams
#210
jerryzh168
closed
4 weeks ago
0
NF4Tensor uses 8 bits of memory
#209
cuichenx
closed
3 weeks ago
7
Enable FSDP Test in CI
#207
msaroufim
closed
4 weeks ago
4
Enable distributed FSDP Tests
#206
msaroufim
closed
1 month ago
2
[NF4][FSDP2] DTensor + fused adam on cpu
#205
weifengpy
closed
3 weeks ago
0
[NF4][FSDP2] avoid peaking GPU memory when constructing NF4 tensors
#204
weifengpy
opened
1 month ago
1
[NF4][FSDP2]: enable multi-gpu CI
#202
weifengpy
opened
1 month ago
0
quant primitives: always set min val for scale
#201
jerryzh168
closed
4 weeks ago
0
Fix README links
#200
jcaip
closed
1 month ago
0
Fix FQ mask in 8da4w QAT
#199
andrewor14
closed
4 weeks ago
3
Previous
Next