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

7
vllm/model_executor/models/dbrx.py
View File

@@ -5,6 +5,9 @@ import torch
import torch.nn as nn
from vllm.attention import Attention, AttentionMetadata
from vllm.distributed import (get_tensor_model_parallel_rank,
get_tensor_model_parallel_world_size,
tensor_model_parallel_all_reduce)
from vllm.model_executor.layers.fused_moe import fused_moe
from vllm.model_executor.layers.linear import (LinearMethodBase,
QKVParallelLinear,
@@ -15,10 +18,6 @@ from vllm.model_executor.layers.rotary_embedding import get_rope
from vllm.model_executor.layers.sampler import Sampler
from vllm.model_executor.layers.vocab_parallel_embedding import (
DEFAULT_VOCAB_PADDING_SIZE, ParallelLMHead, VocabParallelEmbedding)
from vllm.model_executor.parallel_utils.communication_op import (
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.sampling_metadata import SamplingMetadata
from vllm.model_executor.utils import set_weight_attrs
from vllm.model_executor.weight_utils import (default_weight_loader,