A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 16, 08:15 UTC
[V1] Fix Configs (#9971)
Robert Shaw committed
Nov 3, 2024 at 19:24 UTC
91c9ebbb1bfc39e98aa2bd444b9569e5f2f92c9e
1 file changed
+3
-2
vllm/v1/executor/gpu_executor.py
+3
-2
@@ -1,7 +1,7 @@
1
import os
2
from typing import Optional, Tuple
3
4
-from vllm.config import EngineConfig
4
+from vllm.config import VllmConfig
5
from vllm.logger import init_logger
6
from vllm.utils import get_distributed_init_method, get_ip, get_open_port
7
from vllm.v1.outputs import ModelRunnerOutput
@@ -12,7 +12,8 @@ logger = init_logger(__name__)
12
13
class GPUExecutor:
14
15
- def __init__(self, vllm_config: EngineConfig) -> None:
15
+ def __init__(self, vllm_config: VllmConfig) -> None:
16
+ self.vllm_config = vllm_config
17
self.model_config = vllm_config.model_config
18
self.cache_config = vllm_config.cache_config
19
self.lora_config = vllm_config.lora_config