Usage Stats Collection (#2852)
This commit is contained in:
@@ -10,6 +10,7 @@ from vllm.lora.request import LoRARequest
|
||||
from vllm.outputs import RequestOutput
|
||||
from vllm.sampling_params import SamplingParams
|
||||
from vllm.sequence import MultiModalData
|
||||
from vllm.usage.usage_lib import UsageContext
|
||||
from vllm.utils import Counter
|
||||
|
||||
|
||||
@@ -108,7 +109,8 @@ class LLM:
|
||||
disable_custom_all_reduce=disable_custom_all_reduce,
|
||||
**kwargs,
|
||||
)
|
||||
self.llm_engine = LLMEngine.from_engine_args(engine_args)
|
||||
self.llm_engine = LLMEngine.from_engine_args(
|
||||
engine_args, usage_context=UsageContext.LLM_CLASS)
|
||||
self.request_counter = Counter()
|
||||
|
||||
def get_tokenizer(
|
||||
|
||||
Reference in New Issue
Block a user