[Frontend] Port Step-3.5 parsers to the streaming parser engine (#59321)

Signed-off-by: sfeng33 <4florafeng@gmail.com>
This commit is contained in:
Flora Feng
2026-10-01 22:18:37 +00:00
committed by GitHub
parent ca65eb67d9
commit a5105dba05
12 changed files with 236 additions and 1647 deletions
@@ -24,6 +24,7 @@ from vllm.parser.nemotron_v3 import NemotronV3Parser
from vllm.parser.plamo3 import Plamo3Parser
from vllm.parser.qwen3 import Qwen3Parser
from vllm.parser.seed_oss import SeedOssParser
from vllm.parser.step3p5 import Step3p5Parser
(
DeepSeekV32ParserReasoningAdapter,
@@ -80,6 +81,11 @@ from vllm.parser.seed_oss import SeedOssParser
SeedOssParserToolAdapter,
) = make_adapters(SeedOssParser)
(
Step3p5ParserReasoningAdapter,
Step3p5ParserToolAdapter,
) = make_adapters(Step3p5Parser)
(
Glm47MoeParserReasoningAdapter,
Glm47MoeParserToolAdapter,
+49
View File
@@ -0,0 +1,49 @@
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
"""Step-3.5 / Step-3.7 parser for tool calls and reasoning.
Both models use the Qwen3 XML tool-call grammar and ``<think>`` reasoning
unchanged. Two things differ from :class:`Qwen3Parser`:
* The chat template always prefills ``<think>\\n`` and has no switch to turn
thinking off, so ``enable_thinking`` is ignored.
* The model closes reasoning with ``\\n</think>\\n``, so trailing reasoning
whitespace is stripped.
"""
from __future__ import annotations
import functools
from dataclasses import replace
from typing import TYPE_CHECKING
from vllm.parser.engine.parser_engine_config import ParserEngineConfig
from vllm.parser.qwen3 import CHATML_TURN_BOUNDARIES, Qwen3Parser, qwen3_config
if TYPE_CHECKING:
from vllm.tokenizers import TokenizerLike
from vllm.tool_parsers.abstract_tool_parser import Tool
@functools.cache
def step3p5_config() -> ParserEngineConfig:
return replace(
qwen3_config(
thinking=True,
name="step3p5",
turn_boundary_tokens=CHATML_TURN_BOUNDARIES,
),
strip_trailing_reasoning_whitespace=True,
)
class Step3p5Parser(Qwen3Parser):
def __init__(
self,
tokenizer: TokenizerLike,
tools: list[Tool] | None = None,
**kwargs,
) -> None:
kwargs.setdefault("parser_engine_config", step3p5_config())
super().__init__(tokenizer, tools, **kwargs)
self.thinking_enabled = True