[Quantization] Let ModelOpt MXFP8 layers load pre-processed weights Purpose (#57316)
Signed-off-by: liusy58 <mg21330037@smail.nju.edu.cn> Signed-off-by: Isotr0py <2037008807@qq.com> Co-authored-by: Isotr0py <2037008807@qq.com>
This commit is contained in:
@@ -2482,9 +2482,16 @@ class ModelOptLinearMethod(LinearMethodBase):
|
||||
|
||||
@property
|
||||
def supports_pre_processed_weights(self) -> bool: # type: ignore[override]
|
||||
# TODO(Isotr0py): support fp8/mxfp8 ModelOpt kernels transpose/repack.
|
||||
# TODO(Isotr0py): support fp8 ModelOpt kernels transpose/repack.
|
||||
w = self.spec.weight
|
||||
return isinstance(w, QuantKey) and w.dtype == FP4_DTYPE
|
||||
return isinstance(w, QuantKey) and (
|
||||
w.dtype == FP4_DTYPE
|
||||
or (
|
||||
w == kMxfp8Static
|
||||
and self.kernel is not None
|
||||
and self.kernel.supports_pre_processed_weights
|
||||
)
|
||||
)
|
||||
|
||||
def create_weights(
|
||||
self,
|
||||
@@ -2524,7 +2531,13 @@ class ModelOptLinearMethod(LinearMethodBase):
|
||||
expose_input_quant_key(layer, self.kernel)
|
||||
|
||||
def process_weights_after_loading(self, layer) -> None:
|
||||
if self.spec.weight == kMxfp8Static and getattr(layer, "is_bmm", False):
|
||||
self.kernel = init_mxfp8_linear_kernel(bmm_batch_size=layer.bmm_batch_size)
|
||||
if is_weights_pre_processed():
|
||||
if not self.supports_pre_processed_weights:
|
||||
raise RuntimeError(
|
||||
f"{type(self.kernel).__name__} cannot use pre-processed weights"
|
||||
)
|
||||
return
|
||||
self.fmt.pre_process(layer)
|
||||
self.wkey.process(layer, WEIGHT)
|
||||
@@ -2556,8 +2569,6 @@ class ModelOptLinearMethod(LinearMethodBase):
|
||||
persistent=False,
|
||||
)
|
||||
layer._nvfp4_group_size_for_gather = self.ctx.group_size
|
||||
if self.spec.weight == kMxfp8Static and getattr(layer, "is_bmm", False):
|
||||
self.kernel = init_mxfp8_linear_kernel(bmm_batch_size=layer.bmm_batch_size)
|
||||
self.kernel.process_weights_after_loading(layer)
|
||||
|
||||
def apply(self, layer, x, bias=None):
|
||||
|
||||
Reference in New Issue
Block a user