Skip to content

Drop old quantization flows #3115

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

Merged
merged 11 commits into from
Oct 16, 2025
Merged

Drop old quantization flows #3115

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
11 commits
Select commit Hold shift + click to select a range
56d16dc
drop old quantization flows
namgyu-youn Oct 2, 2025
964b7e4
remove old quantization flows
namgyu-youn Oct 2, 2025
fbb2f2b
inline quantization API in benchmarks
namgyu-youn Oct 2, 2025
04ea5f8
Merge branch 'main' into deprecate-old-apis
namgyu-youn Oct 7, 2025
c09839c
drop old apis
namgyu-youn Oct 7, 2025
39e9d9d
fix pre-commit
namgyu-youn Oct 9, 2025
6ef8503
revert smoothquant implementation
namgyu-youn Oct 10, 2025
4651e63
revert smoothquant test
namgyu-youn Oct 10, 2025
79ee25b
revert unrelated changes
namgyu-youn Oct 13, 2025
df20cb5
drop deprecated apis
namgyu-youn Oct 15, 2025
c81e214
fix duplicate import
namgyu-youn Oct 15, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 3 additions & 19 deletions benchmarks/benchmark_aq.py
View file
Open in desktop
Copy link
Contributor Author

@namgyu-youn namgyu-youn Oct 2, 2025 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

How about inlining _int8wo_api, _int8da_int8w_api, _int4wo_api ? They are used only once across codebase.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yeah I think that's fine if they're only used in benchmarks

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

also cc @jainapurva, can you take a look at the benchmark changes?

Original file line number Diff line number Diff line change
Expand Up @@ -16,10 +16,6 @@
_replace_with_custom_fn_if_matches_filter,
quantize_,
)
from torchao.quantization.subclass import (
Int4WeightOnlyQuantizedLinearWeight,
Int8WeightOnlyQuantizedLinearWeight,
)


def _int8wo_api(mod, **kwargs):
Expand Down Expand Up @@ -117,14 +113,6 @@ def _ref_change_linear_weights_to_woqtensors(model, filter_fn=None, **kwargs):
return _ref_change_linear_weights_to_woqtensors


_ref_change_linear_weights_to_int8_woqtensors = (
_get_ref_change_linear_weights_to_woqtensors(Int8WeightOnlyQuantizedLinearWeight)
)
_ref_change_linear_weights_to_int4_woqtensors = (
_get_ref_change_linear_weights_to_woqtensors(Int4WeightOnlyQuantizedLinearWeight)
)


torch._dynamo.config.cache_size_limit = 50000


Expand Down Expand Up @@ -186,20 +174,16 @@ def _bench_quantized_tensor_subclass_perf(api, ref_api, M, N, K, kwargs=None):

for M, N, K in all_shapes:
_bench_quantized_tensor_subclass_perf(
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Temporarily updated to use new APIs 2 times to fix CI, but maybe we can update _bench_quantized_tensor_subclass_perf to compare only original vs. new quantization flows?

_int8da_int8w_api, _ref_change_linear_weights_to_int8_dqtensors, M, N, K
_int8da_int8w_api, _int8da_int8w_api, M, N, K
)

print("_int8wo_api")

for M, N, K in all_shapes:
_bench_quantized_tensor_subclass_perf(
_int8wo_api, _ref_change_linear_weights_to_int8_woqtensors, M, N, K
)
_bench_quantized_tensor_subclass_perf(_int8wo_api, _int8wo_api, M, N, K)

print("_int4wo_api")
kwargs = {"groupsize": 32, "version": 1}

for M, N, K in all_shapes:
_bench_quantized_tensor_subclass_perf(
_int4wo_api, _ref_change_linear_weights_to_int4_woqtensors, M, N, K, kwargs
)
_bench_quantized_tensor_subclass_perf(_int4wo_api, _int4wo_api, M, N, K, kwargs)
Loading
Loading