Skip to content

Issues: pytorch/ao

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

122 Open 77 Closed
122 Open 77 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

installation fails on ARM64 Linux (aka Raspberry Pi 5)
#913 opened Sep 21, 2024 by sunshinesfbay
More fine-grained documentation needed for torchao.autoquant()
#907 opened Sep 19, 2024 by suvadityamuk
@HDCharles @suvadityamuk
3
ValueError: ('Unsupported kind: ', 'FRAGMENT')
#900 opened Sep 17, 2024 by loretoparisi
1
run_vit_b_quant.py runs slower than run_bit_b.py
#898 opened Sep 17, 2024 by jerryzh168
@IvanKobzarev
7
[wip] unification of torchao.float8 with the rest of torchao
#894 opened Sep 16, 2024 by vkuzo
1
we should ensure activation checkpointing with Float8Linear behaves optimally
#893 opened Sep 16, 2024 by vkuzo
1
float8 training with rowwise scaling
#889 opened Sep 16, 2024 by vkuzo
@lw
9
[CI] Add CI test for PyTorch 2.5.0rc
#888 opened Sep 15, 2024 by gau-nernst
Out Of Memory Error
#879 opened Sep 12, 2024 by bianxuxuxu
2
[torchao]NotImplementedError: AffineQuantizedTensor dispatch: attempting to run unimplemented operator/function: aten.permute.default
#890 opened Sep 12, 2024 by yanbing-j
@IvanKobzarev
8
Expected Tensor argument scales to have dtype torch.bfloat16, but got torch.float32 instead
#876 opened Sep 11, 2024 by agunapal
1
Unify scaled INT8 matmul
#862 opened Sep 10, 2024 by gau-nernst
1
1
Add MyPy support enhancement New feature or request
#856 opened Sep 9, 2024 by drisspg
Create report for Autoquant autoquant
#855 opened Sep 9, 2024 by drisspg
Add Autoquant Cache autoquant
#854 opened Sep 9, 2024 by drisspg
1
Update Eval scripts enhancement New feature or request
#853 opened Sep 9, 2024 by drisspg
1
int4_weight_only dequant does not restore original shape
#848 opened Sep 9, 2024 by jerryzh168
AO dtype composability tracker
#844 opened Sep 8, 2024 by msaroufim
2 tasks
1
[Quantization + FSDP] Support quantize_() for DTensor
#803 opened Sep 4, 2024 by gau-nernst
1
NotImplementedError: aten.linear.default not implemented when using MXTensor mx
#796 opened Sep 3, 2024 by Ali-Flt
1
7
placeholder for float8 README.md images
#793 opened Sep 2, 2024 by vkuzo
1
'FakeTensor' object has no attribute 'layout_type'
#789 opened Sep 2, 2024 by WAS-PlaiLabs
4
[Question] Is there any plans to support fp8 batched matmul (_scaled_bmm) float8
#781 opened Aug 30, 2024 by leeeizhang
How does this work with ONNX export and quantization? export inference
#777 opened Aug 29, 2024 by ogencoglu
1
All code snippets on README should be self contained or link to single file runnable code
#765 opened Aug 27, 2024 by msaroufim
2 tasks
ProTip! What’s not been updated in a month: updated:<2024-08-21.