Skip to content

[MXFP4] Support MXFp4 Format #439

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

Draft
wants to merge 4 commits into
base: main
Choose a base branch
from
Draft

[MXFP4] Support MXFp4 Format #439

Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
ae8849e
add format
dsikka Aug 27, 2025
3150cea
update
dsikka Aug 28, 2025
4c11750
update
dsikka Aug 28, 2025
7e4c47d
update
dsikka Aug 28, 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
33 changes: 33 additions & 0 deletions src/compressed_tensors/quantization/quant_scheme.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -162,6 +162,37 @@ def is_preset_scheme(name: str) -> bool:
),
)

MXFP4A16 = dict(
weights=QuantizationArgs(
num_bits=4,
type=QuantizationType.FLOAT,
strategy=QuantizationStrategy.GROUP,
symmetric=True,
dynamic=False,
group_size=32,
)
)

MXFP4 = dict(
weights=QuantizationArgs(
num_bits=4,
type=QuantizationType.FLOAT,
strategy=QuantizationStrategy.GROUP,
symmetric=True,
dynamic=False,
group_size=32,
),
input_activations=QuantizationArgs(
num_bits=4,
type=QuantizationType.FLOAT,
strategy=QuantizationStrategy.GROUP,
dynamic=True,
symmetric=True,
group_size=32,
),
)


# 8 bit integer weights and 8 bit activations quantization
INT8_W8A8 = dict(
weights=QuantizationArgs(
Expand Down Expand Up @@ -313,4 +344,6 @@ def is_preset_scheme(name: str) -> bool:
"FP8_BLOCK": FP8_BLOCK,
"NVFP4A16": NVFP4A16,
"NVFP4": NVFP4,
"MXFP4": MXFP4,
"MXFP4A16": MXFP4A16,
}