[Frontend] Use new Renderer for Completions and Tokenize API (#32863)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
This commit is contained in:
Cyrus Leung
2026-01-31 04:51:15 -08:00
committed by GitHub
parent 8980001c93
commit f0a1c8453a
64 changed files with 2116 additions and 2003 deletions
+26 -4
View File
@@ -4,10 +4,9 @@
import time
from typing import Any, TypeAlias
from pydantic import (
Field,
)
from pydantic import Field
from vllm.config import ModelConfig
from vllm.entrypoints.openai.engine.protocol import OpenAIBaseModel, UsageInfo
from vllm.entrypoints.pooling.base.protocol import (
ChatRequestMixin,
@@ -15,13 +14,24 @@ from vllm.entrypoints.pooling.base.protocol import (
CompletionRequestMixin,
PoolingBasicRequestMixin,
)
from vllm.renderers import TokenizeParams
from vllm.utils import random_uuid
class ClassificationCompletionRequest(
PoolingBasicRequestMixin, CompletionRequestMixin, ClassifyRequestMixin
):
pass
def build_tok_params(self, model_config: ModelConfig) -> TokenizeParams:
encoder_config = model_config.encoder_config or {}
return TokenizeParams(
max_total_tokens=model_config.max_model_len,
max_output_tokens=0,
truncate_prompt_tokens=self.truncate_prompt_tokens,
do_lower_case=encoder_config.get("do_lower_case", False),
add_special_tokens=self.add_special_tokens,
max_total_tokens_param="max_model_len",
)
class ClassificationChatRequest(
@@ -33,6 +43,18 @@ class ClassificationChatRequest(
description=("Additional kwargs to pass to the HF processor."),
)
def build_tok_params(self, model_config: ModelConfig) -> TokenizeParams:
encoder_config = model_config.encoder_config or {}
return TokenizeParams(
max_total_tokens=model_config.max_model_len,
max_output_tokens=0,
truncate_prompt_tokens=self.truncate_prompt_tokens,
do_lower_case=encoder_config.get("do_lower_case", False),
add_special_tokens=self.add_special_tokens,
max_total_tokens_param="max_model_len",
)
ClassificationRequest: TypeAlias = (
ClassificationCompletionRequest | ClassificationChatRequest