[Online quantization] Add online MXFP4 quantization support (#49347)

Signed-off-by: Felix Marty <Felix.Marty@amd.com>
This commit is contained in:
fxmarty-amd
2026-08-07 12:39:46 -07:00
committed by GitHub
parent a671679e9f
commit 3518110f2f
18 changed files with 1423 additions and 35 deletions
+5
View File
@@ -17,6 +17,7 @@ from vllm.model_executor.layers.quantization.utils.quant_utils import (
kFp8StaticTensorSym,
kInt8StaticChannelSym,
kMxfp4Dynamic,
kMxfp4Static,
kMxfp8Dynamic,
kNvfp4Static,
)
@@ -131,6 +132,10 @@ _ONLINE_SHORTHANDS: dict[str, QuantizationConfigArgs] = {
linear=QuantSpec(weight=kMxfp8Dynamic),
moe=QuantSpec(weight=kMxfp8Dynamic),
),
"mxfp4": QuantizationConfigArgs(
linear=QuantSpec(weight=kMxfp4Static),
moe=QuantSpec(weight=kMxfp4Static),
),
# INT8 weight-only on MoE; linear stays unquantized (no `linear` field).
"int8_per_channel_weight_only": QuantizationConfigArgs(
moe=QuantSpec(weight=kInt8StaticChannelSym),