[Core] Refactor model loading code (#4097)

This commit is contained in:
Antoni Baum
2024-04-16 11:34:39 -07:00
committed by GitHub
parent 05434764cd
commit 69e1d2fb69
67 changed files with 1054 additions and 963 deletions
+5 -5
View File
@@ -1,9 +1,9 @@
from abc import ABC, abstractmethod
from typing import Dict, List, Optional, Set, Tuple
from vllm.config import (CacheConfig, DeviceConfig, LoRAConfig, ModelConfig,
ParallelConfig, SchedulerConfig, SpeculativeConfig,
TensorizerConfig, VisionLanguageConfig)
from vllm.config import (CacheConfig, DeviceConfig, LoadConfig, LoRAConfig,
ModelConfig, ParallelConfig, SchedulerConfig,
SpeculativeConfig, VisionLanguageConfig)
from vllm.lora.request import LoRARequest
from vllm.sequence import SamplerOutput, SequenceGroupMetadata
@@ -23,20 +23,20 @@ class ExecutorBase(ABC):
parallel_config: ParallelConfig,
scheduler_config: SchedulerConfig,
device_config: DeviceConfig,
load_config: LoadConfig,
lora_config: Optional[LoRAConfig],
vision_language_config: Optional[VisionLanguageConfig],
speculative_config: Optional[SpeculativeConfig],
tensorizer_config: Optional[TensorizerConfig],
) -> None:
self.model_config = model_config
self.cache_config = cache_config
self.lora_config = lora_config
self.load_config = load_config
self.parallel_config = parallel_config
self.scheduler_config = scheduler_config
self.device_config = device_config
self.vision_language_config = vision_language_config
self.speculative_config = speculative_config
self.tensorizer_config = tensorizer_config
self._init_executor()