biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Core][Refactor] move parallel_utils into vllm/distributed (#3950)

Browse Source 
[WIP][Core][Refactor] move vllm/model_executor/parallel_utils into vllm/distributed and vllm/device_communicators (#3950)
This commit is contained in:
youkaichao
2024-04-10 15:33:30 -07:00
committed by GitHub
parent 934d3662f7
commit 63e7176f26
52 changed files with 111 additions and 141 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
vllm/model_executor/layers/linear.py
View File

@@ -5,13 +5,12 @@ import torch
import torch.nn.functional as F
from torch.nn.parameter import Parameter
from vllm.distributed import (divide, get_tensor_model_parallel_rank,
get_tensor_model_parallel_world_size,
split_tensor_along_last_dim,
tensor_model_parallel_all_gather,
tensor_model_parallel_all_reduce)
from vllm.logger import init_logger
from vllm.model_executor.parallel_utils.communication_op import (
tensor_model_parallel_all_gather, tensor_model_parallel_all_reduce)
from vllm.model_executor.parallel_utils.parallel_state import (
get_tensor_model_parallel_rank, get_tensor_model_parallel_world_size)
from vllm.model_executor.parallel_utils.utils import (
divide, split_tensor_along_last_dim)
from vllm.model_executor.utils import set_weight_attrs
logger = init_logger(__name__)