[Misc] Add indirection layer for custom ops (#3913)

This commit is contained in:
Kunshang Ji
2024-04-10 20:26:07 -07:00
committed by GitHub
parent e42df7227d
commit e9da5a40c6
14 changed files with 224 additions and 32 deletions
+1 -1
View File
@@ -6,7 +6,7 @@ import torch
import torch.nn as nn
import torch.nn.functional as F
from vllm._C import ops
from vllm import _custom_ops as ops
from vllm.distributed import (divide, get_tensor_model_parallel_rank,
get_tensor_model_parallel_world_size)
from vllm.model_executor.layers.quantization import QuantizationConfig
@@ -8,7 +8,7 @@ import torch
import triton
import triton.language as tl
from vllm._C import ops
from vllm import _custom_ops as ops
from vllm.logger import init_logger
from vllm.utils import is_hip
+1 -1
View File
@@ -4,7 +4,7 @@ from typing import Optional, Tuple, Union
import torch
import torch.nn as nn
from vllm._C import ops
from vllm import _custom_ops as ops
class RMSNorm(nn.Module):
@@ -3,7 +3,7 @@ from typing import Any, Dict, List, Optional
import torch
from torch.nn.parameter import Parameter
from vllm._C import ops
from vllm import _custom_ops as ops
from vllm.model_executor.layers.linear import (LinearMethodBase,
set_weight_attrs)
from vllm.model_executor.layers.quantization.base_config import (
@@ -6,7 +6,7 @@ from typing import Any, Dict, List, Optional
import torch
from torch.nn.parameter import Parameter
from vllm._C import ops
from vllm import _custom_ops as ops
from vllm.model_executor.layers.linear import (LinearMethodBase,
set_weight_attrs)
from vllm.model_executor.layers.quantization.base_config import (
@@ -3,7 +3,7 @@ from typing import Any, Dict, List, Optional
import torch
from torch.nn.parameter import Parameter
from vllm._C import ops
from vllm import _custom_ops as ops
from vllm.model_executor.layers.linear import (LinearMethodBase,
set_weight_attrs)
from vllm.model_executor.layers.quantization.base_config import (
@@ -3,7 +3,7 @@ from typing import Any, Dict, List, Optional
import torch
from torch.nn.parameter import Parameter
from vllm._C import ops
from vllm import _custom_ops as ops
from vllm.model_executor.layers.linear import (LinearMethodBase,
set_weight_attrs)
from vllm.model_executor.layers.quantization.base_config import (
@@ -27,7 +27,7 @@ from typing import Any, Dict, List, Optional, Tuple, Union
import torch
import torch.nn as nn
from vllm._C import ops
from vllm import _custom_ops as ops
def _rotate_neox(x: torch.Tensor) -> torch.Tensor: