mirror of
https://github.com/PaddlePaddle/FastDeploy.git
synced 2025-10-04 16:22:57 +08:00
* 延迟 import Config * support chunked_prefill * support chunked_prefill
This commit is contained in:
@@ -18,7 +18,6 @@ from abc import abstractmethod
|
||||
|
||||
import paddle
|
||||
from paddle import nn
|
||||
from paddle.base.core import Config
|
||||
from paddleformers.utils.log import logger
|
||||
|
||||
try:
|
||||
@@ -103,6 +102,8 @@ class DeepEPEngine:
|
||||
|
||||
self.deepep_engine = None
|
||||
|
||||
from paddle.base.core import Config
|
||||
|
||||
self.ep_config = Config(24, 6, 256)
|
||||
self.num_max_dispatch_tokens_per_rank = num_max_dispatch_tokens_per_rank
|
||||
|
||||
|
Reference in New Issue
Block a user