Roll back to use vllm custom allreduce (#3006)

2025-01-20 04:03:15 -08:00
parent dc1881326f
commit 89cd923581
10 changed files with 18 additions and 65 deletions
@@ -5,9 +5,9 @@ from typing import Optional
 import torch
 import torch.nn as nn
 from einops import rearrange, repeat
-from vllm.distributed import parallel_state
-from vllm.distributed import utils as dist_utils

+from sglang.srt.distributed import parallel_state
+from sglang.srt.distributed import utils as dist_utils
 from sglang.srt.layers.attention.triton_ops.prefill_attention import (
    context_attention_fwd,
 )