51 lines
1.3 KiB
Python
51 lines
1.3 KiB
Python
from packaging.version import parse as parse_version
|
|
|
|
try:
|
|
import triton
|
|
|
|
TRITON_AVAILABLE = True
|
|
except ImportError:
|
|
TRITON_AVAILABLE = False
|
|
|
|
try:
|
|
import autogptq_cuda
|
|
|
|
AUTOGPTQ_CUDA_AVAILABLE = True
|
|
except:
|
|
AUTOGPTQ_CUDA_AVAILABLE = False
|
|
|
|
|
|
def dynamically_import_QuantLinear(use_triton: bool, desc_act: bool, group_size: int, bits: int, disable_exllama: bool = False):
|
|
if use_triton:
|
|
from ..nn_modules.qlinear.qlinear_triton import QuantLinear
|
|
else:
|
|
if bits == 4 and not disable_exllama:
|
|
from ..nn_modules.qlinear.qlinear_exllama import QuantLinear
|
|
elif not desc_act or group_size == -1:
|
|
from ..nn_modules.qlinear.qlinear_cuda_old import QuantLinear
|
|
else:
|
|
from ..nn_modules.qlinear.qlinear_cuda import QuantLinear
|
|
|
|
return QuantLinear
|
|
|
|
|
|
def compare_transformers_version(
|
|
version: str = "v4.28.0",
|
|
op: str = "eq"
|
|
):
|
|
assert op in ["eq", "lt", "le", "gt", "ge"]
|
|
|
|
from transformers import __version__
|
|
|
|
return getattr(parse_version(__version__), f"__{op}__")(parse_version(version))
|
|
|
|
|
|
def compare_pytorch_version(
|
|
version: str = "v2.0.0",
|
|
op: str = "eq"
|
|
):
|
|
assert op in ["eq", "lt", "le", "gt", "ge"]
|
|
|
|
from torch import __version__
|
|
|
|
return getattr(parse_version(__version__), f"__{op}__")(parse_version(version))
|