vllm.parser.inkling
¶
Inkling parser: typed content blocks parsed by a single state machine.
Inkling output format (every marker is a dedicated special token)::
<|message_model|><|content_thinking|>...reasoning...<|end_message|>
<|message_model|><|content_text|>...visible text...<|end_message|>
<|message_model|><|content_invoke_tool_json|>
{"name":"get_weather","args":{"city":"SF"}}<|end_message|>
Blocks are self-describing and may repeat in any order; sampling may
also end a block with the standalone <|content_model_end_sampling|>
token. The tool-call payload is a single JSON object whose name is
extracted by the engine's name-from-args path and whose args object
is carved out of the wrapper by _inkling_arg_converter.
Note the terminal labels: THINK_START/THINK_END are what the
engine keys its reasoning plumbing on (is_reasoning_end,
count_reasoning_tokens, initial-state seeding), so <|end_message|>
is labelled THINK_END here even though it ends every block kind —
the transition table, not the label, carries the semantics.
Classes:
InklingParser
¶
Bases: ParserEngine
Methods:
-
adjust_initial_state_from_prompt–Seed the initial parsing state from the prompt tail.
-
adjust_request–Frame an Inkling2 JSON response format with the block grammar.
Source code in vllm/parser/inkling.py
256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 | |
_inkling2_is_reasoning_end(input_ids)
¶
Whether the latest block is a text block of a model message.
A prompt that ends between messages has not ended reasoning: the model may open a thinking block.
Source code in vllm/parser/inkling.py
adjust_initial_state_from_prompt(prompt_token_ids)
¶
Seed the initial parsing state from the prompt tail.
Mirrors the Rust parser's initialize(): scanning the prompt
backwards, the last relevant special token decides whether generation
continues inside a thinking block, a text block, a model message
header, or between blocks.
Source code in vllm/parser/inkling.py
adjust_request(request)
¶
Frame an Inkling2 JSON response format with the block grammar.
As in the reference, any number of free thinking blocks precede the final text block, whose payload alone must match the JSON constraint; the API then returns that payload as the content.
Source code in vllm/parser/inkling.py
_inkling_arg_converter(raw_args, partial)
¶
Carve the args object out of the tool-call JSON wrapper.
Why a converter at all: the engine's tool_args_json machinery
treats the entire TOOL_ARGS text as the tool arguments, but Inkling's
payload is the {"name":...,"args":{...}} wrapper — without a
converter, _compute_arg_delta streams the wrapper verbatim into
the OpenAI arguments field (converter is None -> raw delta,
unconditionally; stream_arg_deltas=False does not stop it).
extract_json_member_object returns a prefix-stable verbatim
substring, which the engine's argument-delta diffing requires.