Files
project_0920/ixformer_inference_functions_init.py

177 lines
3.7 KiB
Python
Raw Normal View History

try:
from .act_and_mul import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .act_bias_mm import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .add import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .bert import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .bnb_dequant import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .bnb_double_quant import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .bnb_mm_dequant import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .bnb_qgemm import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .bnb_quant import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .bnb_rowcol_absmax import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .conv2d import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .cross_entropy_loss import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .flash_attn import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .flash_attn_lib import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .fused_rope import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .gemv import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .groupnorm import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .i8w8o32 import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .layernorm import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .lightllm import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .linalg import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .linear import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .lmdeploy import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .marlin import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .matmul import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .mla_fused import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .mm import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .moe import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .overlap_comm import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .paged_attention import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .quantized_linear import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .residual_bias import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .rms_norm import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .scaled_dot_product_attention import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .smoothquant import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .softmax import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .store_kv_cache import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .t5 import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .tgi import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .vllm import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .w8a8 import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .w8a16 import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .wi4a16 import *
except (ImportError, ModuleNotFoundError):
pass
try:
from .wui4a16 import *
except (ImportError, ModuleNotFoundError):
pass