[V0 Deprecation] Remove Prompt Adapters (#20588)
Signed-off-by: mgoin <mgoin64@gmail.com>
This commit is contained in:
@@ -27,7 +27,6 @@ from vllm.inputs.data import TokensPrompt
|
||||
from vllm.logger import init_logger
|
||||
from vllm.lora.request import LoRARequest
|
||||
from vllm.outputs import PoolingRequestOutput, ScoringRequestOutput
|
||||
from vllm.prompt_adapter.request import PromptAdapterRequest
|
||||
from vllm.transformers_utils.tokenizer import AnyTokenizer, MistralTokenizer
|
||||
from vllm.utils import make_async, merge_async_iterators
|
||||
|
||||
@@ -58,8 +57,6 @@ class ServingScores(OpenAIServing):
|
||||
request_id: str,
|
||||
tokenization_kwargs: Optional[dict[str, Any]] = None,
|
||||
lora_request: Optional[Union[LoRARequest, None]] = None,
|
||||
prompt_adapter_request: Optional[Union[PromptAdapterRequest,
|
||||
None]] = None,
|
||||
trace_headers: Optional[Mapping[str, str]] = None,
|
||||
) -> Union[list[PoolingRequestOutput], ErrorResponse]:
|
||||
input_texts = texts_1 + texts_2
|
||||
@@ -100,8 +97,7 @@ class ServingScores(OpenAIServing):
|
||||
self._log_inputs(request_id_item,
|
||||
input_texts[i],
|
||||
params=pooling_params,
|
||||
lora_request=lora_request,
|
||||
prompt_adapter_request=prompt_adapter_request)
|
||||
lora_request=lora_request)
|
||||
|
||||
generators.append(
|
||||
self.engine_client.encode(
|
||||
@@ -176,8 +172,6 @@ class ServingScores(OpenAIServing):
|
||||
request_id: str,
|
||||
tokenization_kwargs: Optional[dict[str, Any]] = None,
|
||||
lora_request: Optional[Union[LoRARequest, None]] = None,
|
||||
prompt_adapter_request: Optional[Union[PromptAdapterRequest,
|
||||
None]] = None,
|
||||
trace_headers: Optional[Mapping[str, str]] = None,
|
||||
) -> Union[list[PoolingRequestOutput], ErrorResponse]:
|
||||
request_prompts: list[str] = []
|
||||
@@ -261,8 +255,7 @@ class ServingScores(OpenAIServing):
|
||||
self._log_inputs(request_id_item,
|
||||
request_prompts[i],
|
||||
params=pooling_params,
|
||||
lora_request=lora_request,
|
||||
prompt_adapter_request=prompt_adapter_request)
|
||||
lora_request=lora_request)
|
||||
|
||||
generator = self.engine_client.encode(
|
||||
engine_prompt,
|
||||
@@ -295,14 +288,7 @@ class ServingScores(OpenAIServing):
|
||||
raw_request: Optional[Request] = None,
|
||||
truncate_prompt_tokens: Optional[int] = None,
|
||||
) -> Union[list[PoolingRequestOutput], ErrorResponse]:
|
||||
(
|
||||
lora_request,
|
||||
prompt_adapter_request,
|
||||
) = self._maybe_get_adapters(request)
|
||||
|
||||
if prompt_adapter_request is not None:
|
||||
raise NotImplementedError("Prompt adapter is not supported "
|
||||
"for scoring models")
|
||||
lora_request = self._maybe_get_adapters(request)
|
||||
|
||||
tokenizer = await self.engine_client.get_tokenizer(lora_request)
|
||||
|
||||
@@ -340,7 +326,6 @@ class ServingScores(OpenAIServing):
|
||||
request_id=request_id,
|
||||
tokenization_kwargs=tokenization_kwargs,
|
||||
lora_request=lora_request,
|
||||
prompt_adapter_request=prompt_adapter_request,
|
||||
trace_headers=trace_headers)
|
||||
|
||||
else:
|
||||
@@ -352,7 +337,6 @@ class ServingScores(OpenAIServing):
|
||||
request_id=request_id,
|
||||
tokenization_kwargs=tokenization_kwargs,
|
||||
lora_request=lora_request,
|
||||
prompt_adapter_request=prompt_adapter_request,
|
||||
trace_headers=trace_headers)
|
||||
|
||||
async def create_score(
|
||||
|
||||
Reference in New Issue
Block a user