[ROCm][Refactor] Enable AWQMarlinConfig on ROCm to use choose_mp_linear_kernel (#36505)
Signed-off-by: Matthias Gehre <matthias.gehre@amd.com> Co-authored-by: Michael Goin <mgoin64@gmail.com>
This commit is contained in:
co-authored by
Michael Goin
parent
d3fe857135
commit
410d300893
@@ -46,14 +46,15 @@ def query_marlin_supported_quant_types(
|
||||
if current_platform.is_cpu():
|
||||
return _query_cpu_marlin_supported_quant_types(has_zp, include_fp_type)
|
||||
|
||||
if device_capability is None:
|
||||
capability_tuple = current_platform.get_device_capability()
|
||||
device_capability = (
|
||||
-1 if capability_tuple is None else capability_tuple.to_int()
|
||||
)
|
||||
if not current_platform.is_rocm():
|
||||
if device_capability is None:
|
||||
capability_tuple = current_platform.get_device_capability()
|
||||
device_capability = (
|
||||
-1 if capability_tuple is None else capability_tuple.to_int()
|
||||
)
|
||||
|
||||
if device_capability < 75:
|
||||
return []
|
||||
if device_capability < 75:
|
||||
return []
|
||||
|
||||
# - has_zp is True: return quant_types that has zero points
|
||||
# - has_zp is False: return quant_types that has not zero points
|
||||
@@ -210,8 +211,6 @@ def check_marlin_supports_shape(
|
||||
|
||||
|
||||
def check_marlin_supports_layer(layer: LinearBase, group_size: int) -> bool:
|
||||
if current_platform.is_rocm():
|
||||
return False
|
||||
output_size_per_partition = (
|
||||
getattr(layer, "output_size_per_partition", None) or layer.output_size
|
||||
)
|
||||
|
||||
Reference in New Issue
Block a user