Repository navigation
Expand file tree
/
Copy patharg_utils.py
More file actions
1861 lines (1662 loc) · 84.5 KB
/
Copy patharg_utils.py
File metadata and controls
1861 lines (1662 loc) · 84.5 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
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
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
# yapf: disable
import argparse
import copy
import dataclasses
import functools
import json
import sys
from dataclasses import MISSING, dataclass, fields, is_dataclass
from itertools import permutations
from typing import (TYPE_CHECKING, Annotated, Any, Callable, Dict, List,
Literal, Optional, Type, TypeVar, Union, cast, get_args,
get_origin)
import huggingface_hub
import regex as re
import torch
from pydantic import TypeAdapter, ValidationError
from typing_extensions import TypeIs, deprecated
import vllm.envs as envs
from vllm.config import (BlockSize, CacheConfig, CacheDType, CompilationConfig,
ConfigFormat, ConfigType, ConvertOption,
DecodingConfig, DetailedTraceModules, Device,
DeviceConfig, DistributedExecutorBackend, EPLBConfig,
GuidedDecodingBackend, HfOverrides, KVEventsConfig,
KVTransferConfig, LoadConfig, LogprobsMode,
LoRAConfig, MambaDType, MMEncoderTPMode, ModelConfig,
ModelDType, ModelImpl, MultiModalConfig,
ObservabilityConfig, ParallelConfig, PoolerConfig,
PrefixCachingHashAlgo, RunnerOption, SchedulerConfig,
SchedulerPolicy, SpeculativeConfig, TaskOption,
TokenizerMode, VllmConfig, get_attr_docs, get_field)
from vllm.logger import init_logger
from vllm.platforms import CpuArchEnum, current_platform
from vllm.plugins import load_general_plugins
from vllm.ray.lazy_utils import is_ray_initialized
from vllm.reasoning import ReasoningParserManager
from vllm.test_utils import MODEL_WEIGHTS_S3_BUCKET, MODELS_ON_S3
from vllm.transformers_utils.config import get_model_path, is_interleaved
from vllm.transformers_utils.utils import check_gguf_file
from vllm.utils import (STR_DUAL_CHUNK_FLASH_ATTN_VAL, FlexibleArgumentParser,
GiB_bytes, get_ip, is_in_ray_actor)
from vllm.v1.sample.logits_processor import LogitsProcessor
# yapf: enable
if TYPE_CHECKING:
from vllm.executor.executor_base import ExecutorBase
from vllm.model_executor.layers.quantization import QuantizationMethods
from vllm.model_executor.model_loader import LoadFormats
from vllm.usage.usage_lib import UsageContext
else:
ExecutorBase = Any
QuantizationMethods = Any
LoadFormats = Any
UsageContext = Any
logger = init_logger(__name__)
# object is used to allow for special typing forms
T = TypeVar("T")
TypeHint = Union[type[Any], object]
TypeHintT = Union[type[T], object]
def parse_type(return_type: Callable[[str], T]) -> Callable[[str], T]:
def _parse_type(val: str) -> T:
try:
return return_type(val)
except ValueError as e:
raise argparse.ArgumentTypeError(
f"Value {val} cannot be converted to {return_type}.") from e
return _parse_type
def optional_type(
return_type: Callable[[str], T]) -> Callable[[str], Optional[T]]:
def _optional_type(val: str) -> Optional[T]:
if val == "" or val == "None":
return None
return parse_type(return_type)(val)
return _optional_type
def union_dict_and_str(val: str) -> Optional[Union[str, dict[str, str]]]:
if not re.match(r"(?s)^\s*{.*}\s*$", val):
return str(val)
return optional_type(json.loads)(val)
def is_type(type_hint: TypeHint, type: TypeHintT) -> TypeIs[TypeHintT]:
"""Check if the type hint is a specific type."""
return type_hint is type or get_origin(type_hint) is type
def contains_type(type_hints: set[TypeHint], type: TypeHintT) -> bool:
"""Check if the type hints contain a specific type."""
return any(is_type(type_hint, type) for type_hint in type_hints)
def get_type(type_hints: set[TypeHint], type: TypeHintT) -> TypeHintT:
"""Get the specific type from the type hints."""
return next((th for th in type_hints if is_type(th, type)), None)
def literal_to_kwargs(type_hints: set[TypeHint]) -> dict[str, Any]:
"""Get the `type` and `choices` from a `Literal` type hint in `type_hints`.
If `type_hints` also contains `str`, we use `metavar` instead of `choices`.
"""
type_hint = get_type(type_hints, Literal)
options = get_args(type_hint)
option_type = type(options[0])
if not all(isinstance(option, option_type) for option in options):
raise ValueError(
"All options must be of the same type. "
f"Got {options} with types {[type(c) for c in options]}")
kwarg = "metavar" if contains_type(type_hints, str) else "choices"
return {"type": option_type, kwarg: sorted(options)}
def is_not_builtin(type_hint: TypeHint) -> bool:
"""Check if the class is not a built-in type."""
return type_hint.__module__ != "builtins"
def get_type_hints(type_hint: TypeHint) -> set[TypeHint]:
"""Extract type hints from Annotated or Union type hints."""
type_hints: set[TypeHint] = set()
origin = get_origin(type_hint)
args = get_args(type_hint)
if origin is Annotated:
type_hints.update(get_type_hints(args[0]))
elif origin is Union:
for arg in args:
type_hints.update(get_type_hints(arg))
else:
type_hints.add(type_hint)
return type_hints
def is_online_quantization(quantization: Any) -> bool:
return quantization in ["inc"]
@functools.lru_cache(maxsize=30)
def _compute_kwargs(cls: ConfigType) -> dict[str, Any]:
cls_docs = get_attr_docs(cls)
kwargs = {}
for field in fields(cls):
# Get the set of possible types for the field
type_hints: set[TypeHint] = get_type_hints(field.type)
# If the field is a dataclass, we can use the model_validate_json
generator = (th for th in type_hints if is_dataclass(th))
dataclass_cls = next(generator, None)
# Get the default value of the field
if field.default is not MISSING:
default = field.default
elif field.default_factory is not MISSING:
default = field.default_factory()
# Get the help text for the field
name = field.name
help = cls_docs[name].strip()
# Escape % for argparse
help = help.replace("%", "%%")
# Initialise the kwargs dictionary for the field
kwargs[name] = {"default": default, "help": help}
# Set other kwargs based on the type hints
json_tip = ("Should either be a valid JSON string or JSON keys passed "
"individually.")
if dataclass_cls is not None:
def parse_dataclass(val: str, cls=dataclass_cls) -> Any:
try:
return TypeAdapter(cls).validate_json(val)
except ValidationError as e:
raise argparse.ArgumentTypeError(repr(e)) from e
kwargs[name]["type"] = parse_dataclass
kwargs[name]["help"] += f"\n\n{json_tip}"
elif contains_type(type_hints, bool):
# Creates --no-<name> and --<name> flags
kwargs[name]["action"] = argparse.BooleanOptionalAction
elif contains_type(type_hints, Literal):
kwargs[name].update(literal_to_kwargs(type_hints))
elif contains_type(type_hints, tuple):
type_hint = get_type(type_hints, tuple)
types = get_args(type_hint)
tuple_type = types[0]
assert all(t is tuple_type for t in types if t is not Ellipsis), (
"All non-Ellipsis tuple elements must be of the same "
f"type. Got {types}.")
kwargs[name]["type"] = tuple_type
kwargs[name]["nargs"] = "+" if Ellipsis in types else len(types)
elif contains_type(type_hints, list):
type_hint = get_type(type_hints, list)
types = get_args(type_hint)
list_type = types[0]
if get_origin(list_type) is Union:
msg = "List type must contain str if it is a Union."
assert str in get_args(list_type), msg
list_type = str
kwargs[name]["type"] = list_type
kwargs[name]["nargs"] = "+"
elif contains_type(type_hints, int):
kwargs[name]["type"] = int
# Special case for large integers
if name in {"max_model_len", "max_num_batched_tokens"}:
kwargs[name]["type"] = human_readable_int
elif contains_type(type_hints, float):
kwargs[name]["type"] = float
elif (contains_type(type_hints, dict)
and (contains_type(type_hints, str)
or any(is_not_builtin(th) for th in type_hints))):
kwargs[name]["type"] = union_dict_and_str
elif contains_type(type_hints, dict):
kwargs[name]["type"] = parse_type(json.loads)
kwargs[name]["help"] += f"\n\n{json_tip}"
elif (contains_type(type_hints, str)
or any(is_not_builtin(th) for th in type_hints)):
kwargs[name]["type"] = str
else:
raise ValueError(
f"Unsupported type {type_hints} for argument {name}.")
# If the type hint was a sequence of literals, use the helper function
# to update the type and choices
if get_origin(kwargs[name].get("type")) is Literal:
kwargs[name].update(literal_to_kwargs({kwargs[name]["type"]}))
# If None is in type_hints, make the argument optional.
# But not if it's a bool, argparse will handle this better.
if type(None) in type_hints and not contains_type(type_hints, bool):
kwargs[name]["type"] = optional_type(kwargs[name]["type"])
if kwargs[name].get("choices"):
kwargs[name]["choices"].append("None")
return kwargs
def get_kwargs(cls: ConfigType) -> dict[str, Any]:
"""Return argparse kwargs for the given Config dataclass.
The heavy computation is cached via functools.lru_cache, and a deep copy
is returned so callers can mutate the dictionary without affecting the
cached version.
"""
return copy.deepcopy(_compute_kwargs(cls))
@dataclass
class EngineArgs:
"""Arguments for vLLM engine."""
model: str = ModelConfig.model
served_model_name: Optional[Union[
str, List[str]]] = ModelConfig.served_model_name
tokenizer: Optional[str] = ModelConfig.tokenizer
hf_config_path: Optional[str] = ModelConfig.hf_config_path
runner: RunnerOption = ModelConfig.runner
convert: ConvertOption = ModelConfig.convert
task: Optional[TaskOption] = ModelConfig.task
skip_tokenizer_init: bool = ModelConfig.skip_tokenizer_init
enable_prompt_embeds: bool = ModelConfig.enable_prompt_embeds
tokenizer_mode: TokenizerMode = ModelConfig.tokenizer_mode
trust_remote_code: bool = ModelConfig.trust_remote_code
allowed_local_media_path: str = ModelConfig.allowed_local_media_path
download_dir: Optional[str] = LoadConfig.download_dir
load_format: Union[str, LoadFormats] = LoadConfig.load_format
config_format: str = ModelConfig.config_format
dtype: ModelDType = ModelConfig.dtype
kv_cache_dtype: CacheDType = CacheConfig.cache_dtype
seed: Optional[int] = ModelConfig.seed
max_model_len: Optional[int] = ModelConfig.max_model_len
cuda_graph_sizes: list[int] = get_field(SchedulerConfig,
"cuda_graph_sizes")
# Note: Specifying a custom executor backend by passing a class
# is intended for expert use only. The API may change without
# notice.
distributed_executor_backend: Optional[Union[
str, DistributedExecutorBackend,
Type[ExecutorBase]]] = ParallelConfig.distributed_executor_backend
# number of P/D disaggregation (or other disaggregation) workers
pipeline_parallel_size: int = ParallelConfig.pipeline_parallel_size
tensor_parallel_size: int = ParallelConfig.tensor_parallel_size
data_parallel_size: int = ParallelConfig.data_parallel_size
data_parallel_rank: Optional[int] = None
data_parallel_start_rank: Optional[int] = None
data_parallel_size_local: Optional[int] = None
data_parallel_address: Optional[str] = None
data_parallel_rpc_port: Optional[int] = None
data_parallel_hybrid_lb: bool = False
data_parallel_backend: str = ParallelConfig.data_parallel_backend
enable_expert_parallel: bool = ParallelConfig.enable_expert_parallel
eplb_config: EPLBConfig = get_field(ParallelConfig, "eplb_config")
enable_eplb: bool = ParallelConfig.enable_eplb
num_redundant_experts: int = EPLBConfig.num_redundant_experts
eplb_window_size: int = EPLBConfig.window_size
eplb_step_interval: int = EPLBConfig.step_interval
eplb_log_balancedness: bool = EPLBConfig.log_balancedness
max_parallel_loading_workers: Optional[
int] = ParallelConfig.max_parallel_loading_workers
block_size: Optional[BlockSize] = CacheConfig.block_size
enable_prefix_caching: Optional[bool] = CacheConfig.enable_prefix_caching
prefix_caching_hash_algo: PrefixCachingHashAlgo = \
CacheConfig.prefix_caching_hash_algo
disable_sliding_window: bool = ModelConfig.disable_sliding_window
disable_cascade_attn: bool = ModelConfig.disable_cascade_attn
swap_space: float = CacheConfig.swap_space
cpu_offload_gb: float = CacheConfig.cpu_offload_gb
gpu_memory_utilization: float = CacheConfig.gpu_memory_utilization
max_num_batched_tokens: Optional[
int] = SchedulerConfig.max_num_batched_tokens
max_num_partial_prefills: int = SchedulerConfig.max_num_partial_prefills
max_long_partial_prefills: int = SchedulerConfig.max_long_partial_prefills
long_prefill_token_threshold: int = \
SchedulerConfig.long_prefill_token_threshold
max_num_seqs: Optional[int] = SchedulerConfig.max_num_seqs
max_logprobs: int = ModelConfig.max_logprobs
logprobs_mode: LogprobsMode = ModelConfig.logprobs_mode
disable_log_stats: bool = False
revision: Optional[str] = ModelConfig.revision
code_revision: Optional[str] = ModelConfig.code_revision
rope_scaling: dict[str, Any] = get_field(ModelConfig, "rope_scaling")
rope_theta: Optional[float] = ModelConfig.rope_theta
hf_token: Optional[Union[bool, str]] = ModelConfig.hf_token
hf_overrides: HfOverrides = get_field(ModelConfig, "hf_overrides")
tokenizer_revision: Optional[str] = ModelConfig.tokenizer_revision
quantization: Optional[QuantizationMethods] = ModelConfig.quantization
enforce_eager: bool = ModelConfig.enforce_eager
max_seq_len_to_capture: int = ModelConfig.max_seq_len_to_capture
disable_custom_all_reduce: bool = ParallelConfig.disable_custom_all_reduce
limit_mm_per_prompt: dict[str, int] = \
get_field(MultiModalConfig, "limit_per_prompt")
interleave_mm_strings: bool = MultiModalConfig.interleave_mm_strings
media_io_kwargs: dict[str, dict[str,
Any]] = get_field(MultiModalConfig,
"media_io_kwargs")
mm_processor_kwargs: Optional[Dict[str, Any]] = \
MultiModalConfig.mm_processor_kwargs
disable_mm_preprocessor_cache: bool = False # DEPRECATED
mm_processor_cache_gb: int = MultiModalConfig.mm_processor_cache_gb
mm_encoder_tp_mode: MMEncoderTPMode = MultiModalConfig.mm_encoder_tp_mode
skip_mm_profiling: bool = MultiModalConfig.skip_mm_profiling
# LoRA fields
enable_lora: bool = False
enable_lora_bias: bool = LoRAConfig.bias_enabled
max_loras: int = LoRAConfig.max_loras
max_lora_rank: int = LoRAConfig.max_lora_rank
default_mm_loras: Optional[Dict[str, str]] = \
LoRAConfig.default_mm_loras
fully_sharded_loras: bool = LoRAConfig.fully_sharded_loras
max_cpu_loras: Optional[int] = LoRAConfig.max_cpu_loras
lora_dtype: Optional[Union[str, torch.dtype]] = LoRAConfig.lora_dtype
lora_extra_vocab_size: int = LoRAConfig.lora_extra_vocab_size
ray_workers_use_nsight: bool = ParallelConfig.ray_workers_use_nsight
num_gpu_blocks_override: Optional[
int] = CacheConfig.num_gpu_blocks_override
num_lookahead_slots: int = SchedulerConfig.num_lookahead_slots
model_loader_extra_config: dict = \
get_field(LoadConfig, "model_loader_extra_config")
ignore_patterns: Optional[Union[str,
List[str]]] = LoadConfig.ignore_patterns
preemption_mode: Optional[str] = SchedulerConfig.preemption_mode
scheduler_delay_factor: float = SchedulerConfig.delay_factor
enable_chunked_prefill: Optional[
bool] = SchedulerConfig.enable_chunked_prefill
disable_chunked_mm_input: bool = SchedulerConfig.disable_chunked_mm_input
disable_hybrid_kv_cache_manager: bool = (
SchedulerConfig.disable_hybrid_kv_cache_manager)
guided_decoding_backend: GuidedDecodingBackend = DecodingConfig.backend
guided_decoding_disable_fallback: bool = DecodingConfig.disable_fallback
guided_decoding_disable_any_whitespace: bool = \
DecodingConfig.disable_any_whitespace
guided_decoding_disable_additional_properties: bool = \
DecodingConfig.disable_additional_properties
logits_processor_pattern: Optional[
str] = ModelConfig.logits_processor_pattern
speculative_config: Optional[Dict[str, Any]] = None
show_hidden_metrics_for_version: Optional[str] = \
ObservabilityConfig.show_hidden_metrics_for_version
otlp_traces_endpoint: Optional[str] = \
ObservabilityConfig.otlp_traces_endpoint
collect_detailed_traces: Optional[list[DetailedTraceModules]] = \
ObservabilityConfig.collect_detailed_traces
disable_async_output_proc: bool = not ModelConfig.use_async_output_proc
scheduling_policy: SchedulerPolicy = SchedulerConfig.policy
scheduler_cls: Union[str, Type[object]] = SchedulerConfig.scheduler_cls
override_neuron_config: dict[str, Any] = \
get_field(ModelConfig, "override_neuron_config")
override_pooler_config: Optional[Union[dict, PoolerConfig]] = \
ModelConfig.override_pooler_config
compilation_config: CompilationConfig = \
get_field(VllmConfig, "compilation_config")
worker_cls: str = ParallelConfig.worker_cls
worker_extension_cls: str = ParallelConfig.worker_extension_cls
kv_transfer_config: Optional[KVTransferConfig] = None
kv_events_config: Optional[KVEventsConfig] = None
generation_config: str = ModelConfig.generation_config
enable_sleep_mode: bool = ModelConfig.enable_sleep_mode
override_generation_config: dict[str, Any] = \
get_field(ModelConfig, "override_generation_config")
model_impl: str = ModelConfig.model_impl
override_attention_dtype: str = ModelConfig.override_attention_dtype
calculate_kv_scales: bool = CacheConfig.calculate_kv_scales
mamba_cache_dtype: MambaDType = CacheConfig.mamba_cache_dtype
mamba_ssm_cache_dtype: MambaDType = CacheConfig.mamba_ssm_cache_dtype
additional_config: dict[str, Any] = \
get_field(VllmConfig, "additional_config")
reasoning_parser: str = DecodingConfig.reasoning_backend
use_tqdm_on_load: bool = LoadConfig.use_tqdm_on_load
pt_load_map_location: str = LoadConfig.pt_load_map_location
# DEPRECATED
enable_multimodal_encoder_data_parallel: bool = False
logits_processors: Optional[list[Union[
str, type[LogitsProcessor]]]] = ModelConfig.logits_processors
"""Custom logitproc types"""
async_scheduling: bool = SchedulerConfig.async_scheduling
kv_sharing_fast_prefill: bool = \
CacheConfig.kv_sharing_fast_prefill
def __post_init__(self):
# support `EngineArgs(compilation_config={...})`
# without having to manually construct a
# CompilationConfig object
if isinstance(self.compilation_config, dict):
self.compilation_config = CompilationConfig(
**self.compilation_config)
if isinstance(self.eplb_config, dict):
self.eplb_config = EPLBConfig.from_cli(json.dumps(
self.eplb_config))
# Setup plugins
from vllm.plugins import load_general_plugins
load_general_plugins()
# when use hf offline,replace model id to local model path
if huggingface_hub.constants.HF_HUB_OFFLINE:
model_id = self.model
self.model = get_model_path(self.model, self.revision)
logger.info(
"HF_HUB_OFFLINE is True, replace model_id [%s] " \
"to model_path [%s]",model_id, self.model)
@staticmethod
def add_cli_args(parser: FlexibleArgumentParser) -> FlexibleArgumentParser:
"""Shared CLI arguments for vLLM engine."""
# Model arguments
model_kwargs = get_kwargs(ModelConfig)
model_group = parser.add_argument_group(
title="ModelConfig",
description=ModelConfig.__doc__,
)
if not ('serve' in sys.argv[1:] and '--help' in sys.argv[1:]):
model_group.add_argument("--model", **model_kwargs["model"])
model_group.add_argument("--runner", **model_kwargs["runner"])
model_group.add_argument("--convert", **model_kwargs["convert"])
model_group.add_argument("--task",
**model_kwargs["task"],
deprecated=True)
model_group.add_argument("--tokenizer", **model_kwargs["tokenizer"])
model_group.add_argument("--tokenizer-mode",
**model_kwargs["tokenizer_mode"])
model_group.add_argument("--trust-remote-code",
**model_kwargs["trust_remote_code"])
model_group.add_argument("--dtype", **model_kwargs["dtype"])
model_group.add_argument("--seed", **model_kwargs["seed"])
model_group.add_argument("--hf-config-path",
**model_kwargs["hf_config_path"])
model_group.add_argument("--allowed-local-media-path",
**model_kwargs["allowed_local_media_path"])
model_group.add_argument("--revision", **model_kwargs["revision"])
model_group.add_argument("--code-revision",
**model_kwargs["code_revision"])
model_group.add_argument("--rope-scaling",
**model_kwargs["rope_scaling"])
model_group.add_argument("--rope-theta", **model_kwargs["rope_theta"])
model_group.add_argument("--tokenizer-revision",
**model_kwargs["tokenizer_revision"])
model_group.add_argument("--max-model-len",
**model_kwargs["max_model_len"])
model_group.add_argument("--quantization", "-q",
**model_kwargs["quantization"])
model_group.add_argument("--enforce-eager",
**model_kwargs["enforce_eager"])
model_group.add_argument("--max-seq-len-to-capture",
**model_kwargs["max_seq_len_to_capture"])
model_group.add_argument("--max-logprobs",
**model_kwargs["max_logprobs"])
model_group.add_argument("--logprobs-mode",
choices=[f.value for f in LogprobsMode],
**model_kwargs["logprobs_mode"])
model_group.add_argument("--disable-sliding-window",
**model_kwargs["disable_sliding_window"])
model_group.add_argument("--disable-cascade-attn",
**model_kwargs["disable_cascade_attn"])
model_group.add_argument("--skip-tokenizer-init",
**model_kwargs["skip_tokenizer_init"])
model_group.add_argument("--enable-prompt-embeds",
**model_kwargs["enable_prompt_embeds"])
model_group.add_argument("--served-model-name",
**model_kwargs["served_model_name"])
# This one is a special case because it is the
# opposite of ModelConfig.use_async_output_proc
model_group.add_argument(
"--disable-async-output-proc",
action="store_true",
<
10BC8
div>
default=EngineArgs.disable_async_output_proc,
help="Disable async output processing. This may result in "
"lower performance.")
model_group.add_argument("--config-format",
choices=[f.value for f in ConfigFormat],
**model_kwargs["config_format"])
# This one is a special case because it can bool
# or str. TODO: Handle this in get_kwargs
model_group.add_argument("--hf-token",
type=str,
nargs="?",
const=True,
default=model_kwargs["hf_token"]["default"],
help=model_kwargs["hf_token"]["help"])
model_group.add_argument("--hf-overrides",
**model_kwargs["hf_overrides"])
model_group.add_argument("--override-neuron-config",
**model_kwargs["override_neuron_config"])
model_group.add_argument("--override-pooler-config",
**model_kwargs["override_pooler_config"])
model_group.add_argument("--logits-processor-pattern",
**model_kwargs["logits_processor_pattern"])
model_group.add_argument("--generation-config",
**model_kwargs["generation_config"])
model_group.add_argument("--override-generation-config",
**model_kwargs["override_generation_config"])
model_group.add_argument("--enable-sleep-mode",
**model_kwargs["enable_sleep_mode"])
model_group.add_argument("--model-impl",
choices=[f.value for f in ModelImpl],
**model_kwargs["model_impl"])
model_group.add_argument("--override-attention-dtype",
**model_kwargs["override_attention_dtype"])
model_group.add_argument("--logits-processors",
**model_kwargs["logits_processors"])
# Model loading arguments
load_kwargs = get_kwargs(LoadConfig)
load_group = parser.add_argument_group(
title="LoadConfig",
description=LoadConfig.__doc__,
)
load_group.add_argument("--load-format", **load_kwargs["load_format"])
load_group.add_argument("--download-dir",
**load_kwargs["download_dir"])
load_group.add_argument("--model-loader-extra-config",
**load_kwargs["model_loader_extra_config"])
load_group.add_argument("--ignore-patterns",
**load_kwargs["ignore_patterns"])
load_group.add_argument("--use-tqdm-on-load",
**load_kwargs["use_tqdm_on_load"])
load_group.add_argument('--pt-load-map-location',
**load_kwargs["pt_load_map_location"])
# Guided decoding arguments
guided_decoding_kwargs = get_kwargs(DecodingConfig)
guided_decoding_group = parser.add_argument_group(
title="DecodingConfig",
description=DecodingConfig.__doc__,
)
guided_decoding_group.add_argument("--guided-decoding-backend",
**guided_decoding_kwargs["backend"])
guided_decoding_group.add_argument(
"--guided-decoding-disable-fallback",
**guided_decoding_kwargs["disable_fallback"])
guided_decoding_group.add_argument(
"--guided-decoding-disable-any-whitespace",
**guided_decoding_kwargs["disable_any_whitespace"])
guided_decoding_group.add_argument(
"--guided-decoding-disable-additional-properties",
**guided_decoding_kwargs["disable_additional_properties"])
guided_decoding_group.add_argument(
"--reasoning-parser",
# This choices is a special case because it's not static
choices=list(ReasoningParserManager.reasoning_parsers),
**guided_decoding_kwargs["reasoning_backend"])
# Parallel arguments
parallel_kwargs = get_kwargs(ParallelConfig)
parallel_group = parser.add_argument_group(
title="ParallelConfig",
description=ParallelConfig.__doc__,
)
parallel_group.add_argument(
"--distributed-executor-backend",
**parallel_kwargs["distributed_executor_backend"])
parallel_group.add_argument(
"--pipeline-parallel-size", "-pp",
**parallel_kwargs["pipeline_parallel_size"])
parallel_group.add_argument("--tensor-parallel-size", "-tp",
**parallel_kwargs["tensor_parallel_size"])
parallel_group.add_argument("--data-parallel-size", "-dp",
**parallel_kwargs["data_parallel_size"])
parallel_group.add_argument(
'--data-parallel-rank',
'-dpn',
type=int,
help='Data parallel rank of this instance. '
'When set, enables external load balancer mode.')
parallel_group.add_argument('--data-parallel-start-rank',
'-dpr',
type=int,
help='Starting data parallel rank '
'for secondary nodes.')
parallel_group.add_argument('--data-parallel-size-local',
'-dpl',
type=int,
help='Number of data parallel replicas '
'to run on this node.')
parallel_group.add_argument('--data-parallel-address',
'-dpa',
type=str,
help='Address of data parallel cluster '
'head-node.')
parallel_group.add_argument('--data-parallel-rpc-port',
'-dpp',
type=int,
help='Port for data parallel RPC '
'communication.')
parallel_group.add_argument('--data-parallel-backend',
'-dpb',
type=str,
default='mp',
help='Backend for data parallel, either '
'"mp" or "ray".')
parallel_group.add_argument(
"--data-parallel-hybrid-lb",
**parallel_kwargs["data_parallel_hybrid_lb"])
parallel_group.add_argument(
"--enable-expert-parallel",
**parallel_kwargs["enable_expert_parallel"])
parallel_group.add_argument("--enable-eplb",
**parallel_kwargs["enable_eplb"])
parallel_group.add_argument("--eplb-config",
**parallel_kwargs["eplb_config"])
parallel_group.add_argument(
"--num-redundant-experts",
type=int,
help=
"[DEPRECATED] --num-redundant-experts will be removed in v0.12.0.",
deprecated=True)
parallel_group.add_argument(
"--eplb-window-size",
type=int,
help="[DEPRECATED] --eplb-window-size will be removed in v0.12.0.",
deprecated=True)
parallel_group.add_argument(
"--eplb-step-interval",
type=int,
help=
"[DEPRECATED] --eplb-step-interval will be removed in v0.12.0.",
deprecated=True)
parallel_group.add_argument(
"--eplb-log-balancedness",
action=argparse.BooleanOptionalAction,
help=
"[DEPRECATED] --eplb-log-balancedness will be removed in v0.12.0.",
deprecated=True)
parallel_group.add_argument(
"--max-parallel-loading-workers",
**parallel_kwargs["max_parallel_loading_workers"])
parallel_group.add_argument(
"--ray-workers-use-nsight",
**parallel_kwargs["ray_workers_use_nsight"])
parallel_group.add_argument(
"--disable-custom-all-reduce",
**parallel_kwargs["disable_custom_all_reduce"])
parallel_group.add_argument("--worker-cls",
**parallel_kwargs["worker_cls"])
parallel_group.add_argument("--worker-extension-cls",
**parallel_kwargs["worker_extension_cls"])
parallel_group.add_argument(
"--enable-multimodal-encoder-data-parallel",
action="store_true",
deprecated=True)
# KV cache arguments
cache_kwargs = get_kwargs(CacheConfig)
cache_group = parser.add_argument_group(
title="CacheConfig",
description=CacheConfig.__doc__,
)
cache_group.add_argument("--block-size", **cache_kwargs["block_size"])
cache_group.add_argument("--gpu-memory-utilization",
**cache_kwargs["gpu_memory_utilization"])
cache_group.add_argument("--swap-space", **cache_kwargs["swap_space"])
cache_group.add_argument("--kv-cache-dtype",
**cache_kwargs["cache_dtype"])
cache_group.add_argument("--num-gpu-blocks-override",
**cache_kwargs["num_gpu_blocks_override"])
cache_group.add_argument("--enable-prefix-caching",
**cache_kwargs["enable_prefix_caching"])
cache_group.add_argument("--prefix-caching-hash-algo",
**cache_kwargs["prefix_caching_hash_algo"])
cache_group.add_argument("--cpu-offload-gb",
**cache_kwargs["cpu_offload_gb"])
cache_group.add_argument("--calculate-kv-scales",
**cache_kwargs["calculate_kv_scales"])
cache_group.add_argument("--kv-sharing-fast-prefill",
**cache_kwargs["kv_sharing_fast_prefill"])
cache_group.add_argument("--mamba-cache-dtype",
**cache_kwargs["mamba_cache_dtype"])
cache_group.add_argument("--mamba-ssm-cache-dtype",
**cache_kwargs["mamba_ssm_cache_dtype"])
# Multimodal related configs
multimodal_kwargs = get_kwargs(MultiModalConfig)
multimodal_group = parser.add_argument_group(
title="MultiModalConfig",
description=MultiModalConfig.__doc__,
)
multimodal_group.add_argument("--limit-mm-per-prompt",
**multimodal_kwargs["limit_per_prompt"])
multimodal_group.add_argument("--media-io-kwargs",
**multimodal_kwargs["media_io_kwargs"])
multimodal_group.add_argument(
"--mm-processor-kwargs",
**multimodal_kwargs["mm_processor_kwargs"])
multimodal_group.add_argument(
"--mm-processor-cache-gb",
**multimodal_kwargs["mm_processor_cache_gb"])
multimodal_group.add_argument("--disable-mm-preprocessor-cache",
action="store_true",
deprecated=True)
multimodal_group.add_argument(
"--mm-encoder-tp-mode", **multimodal_kwargs["mm_encoder_tp_mode"])
multimodal_group.add_argument(
"--interleave-mm-strings",
**multimodal_kwargs["interleave_mm_strings"])
multimodal_group.add_argument("--skip-mm-profiling",
**multimodal_kwargs["skip_mm_profiling"])
# LoRA related configs
lora_kwargs = get_kwargs(LoRAConfig)
lora_group = parser.add_argument_group(
title="LoRAConfig",
description=LoRAConfig.__doc__,
)
lora_group.add_argument(
"--enable-lora",
action=argparse.BooleanOptionalAction,
help="If True, enable handling of LoRA adapters.")
lora_group.add_argument("--enable-lora-bias",
**lora_kwargs["b
10BC8
ias_enabled"])
lora_group.add_argument("--max-loras", **lora_kwargs["max_loras"])
lora_group.add_argument("--max-lora-rank",
**lora_kwargs["max_lora_rank"])
lora_group.add_argument("--lora-extra-vocab-size",
**lora_kwargs["lora_extra_vocab_size"])
lora_group.add_argument(
"--lora-dtype",
**lora_kwargs["lora_dtype"],
)
lora_group.add_argument("--max-cpu-loras",
**lora_kwargs["max_cpu_loras"])
lora_group.add_argument("--fully-sharded-loras",
**lora_kwargs["fully_sharded_loras"])
lora_group.add_argument("--default-mm-loras",
**lora_kwargs["default_mm_loras"])
# Observability arguments
observability_kwargs = get_kwargs(ObservabilityConfig)
observability_group = parser.add_argument_group(
title="ObservabilityConfig",
description=ObservabilityConfig.__doc__,
)
observability_group.add_argument(
"--show-hidden-metrics-for-version",
**observability_kwargs["show_hidden_metrics_for_version"])
observability_group.add_argument(
"--otlp-traces-endpoint",
**observability_kwargs["otlp_traces_endpoint"])
# TODO: generalise this special case
choices = observability_kwargs["collect_detailed_traces"]["choices"]
metavar = f"{{{','.join(choices)}}}"
observability_kwargs["collect_detailed_traces"]["metavar"] = metavar
observability_kwargs["collect_detailed_traces"]["choices"] += [
",".join(p)
for p in permutations(get_args(DetailedTraceModules), r=2)
]
observability_group.add_argument(
"--collect-detailed-traces",
**observability_kwargs["collect_detailed_traces"])
# Scheduler arguments
scheduler_kwargs = get_kwargs(SchedulerConfig)
scheduler_group = parser.add_argument_group(
title="SchedulerConfig",
description=SchedulerConfig.__doc__,
)
scheduler_group.add_argument(
"--max-num-batched-tokens",
**scheduler_kwargs["max_num_batched_tokens"])
scheduler_group.add_argument("--max-num-seqs",
**scheduler_kwargs["max_num_seqs"])
scheduler_group.add_argument(
"--max-num-partial-prefills",
**scheduler_kwargs["max_num_partial_prefills"])
scheduler_group.add_argument(
"--max-long-partial-prefills",
**scheduler_kwargs["max_long_partial_prefills"])
scheduler_group.add_argument('--cuda-graph-sizes',
**scheduler_kwargs["cuda_graph_sizes"])
scheduler_group.add_argument(
"--long-prefill-token-threshold",
**scheduler_kwargs["long_prefill_token_threshold"])
scheduler_group.add_argument("--num-lookahead-slots",
**scheduler_kwargs["num_lookahead_slots"])
scheduler_group.add_argument("--scheduler-delay-factor",
**scheduler_kwargs["delay_factor"])
scheduler_group.add_argument("--preemption-mode",
**scheduler_kwargs["preemption_mode"])
# multi-step scheduling has been removed; corresponding arguments
# are no longer supported.
scheduler_group.add_argument("--scheduling-policy",
**scheduler_kwargs["policy"])
scheduler_group.add_argument(
"--enable-chunked-prefill",
**scheduler_kwargs["enable_chunked_prefill"])
scheduler_group.add_argument(
"--disable-chunked-mm-input",
**scheduler_kwargs["disable_chunked_mm_input"])
scheduler_group.add_argument("--scheduler-cls",
**scheduler_kwargs["scheduler_cls"])
scheduler_group.add_argument(
"--disable-hybrid-kv-cache-manager",
**scheduler_kwargs["disable_hybrid_kv_cache_manager"])
scheduler_group.add_argument("--async-scheduling",
**scheduler_kwargs["async_scheduling"])
# vLLM arguments
vllm_kwargs = get_kwargs(VllmConfig)
vllm_group = parser.add_argument_group(
title="VllmConfig",
description=VllmConfig.__doc__,
)
# We construct SpeculativeConfig using fields from other configs in
# create_engine_config. So we set the type to a JSON string here to
# delay the Pydantic validation that comes with SpeculativeConfig.
vllm_kwargs["speculative_config"]["type"] = optional_type(json.loads)
vllm_group.add_argument("--speculative-config",
**vllm_kwargs["speculative_config"])
vllm_group.add_argument("--kv-transfer-config",
**vllm_kwargs["kv_transfer_config"])
vllm_group.add_argument('--kv-events-config',
**vllm_kwargs["kv_events_config"])
vllm_group.add_argument("--compilation-config", "-O",
**vllm_kwargs["compilation_config"])
vllm_group.add_argument("--additional-config",
**vllm_kwargs["additional_config"])
# Other arguments
parser.add_argument('--disable-log-stats',
action='store_true',
help='Disable logging statistics.')
return parser
@classmethod
def from_cli_args(cls, args: argparse.Namespace):
# Get the list of attributes of this dataclass.
attrs = [attr.name for attr in dataclasses.fields(cls)]
# Set the attributes from the parsed arguments.
engine_args = cls(**{attr: getattr(args, attr) for attr in attrs})
return engine_args
def create_model_config(self) -> ModelConfig:
# gguf file needs a specific model loader and doesn't use hf_repo
if check_gguf_file(self.model):
self.quantization = self.load_format = "gguf"
# NOTE: This is to allow model loading from S3 in CI
if (not isinstance(self, AsyncEngineArgs) and envs.VLLM_CI_USE_S3
and self.model in MODELS_ON_S3 and self.load_format == "auto"):
self.model = f"{MODEL_WEIGHTS_S3_BUCKET}/{self.model}"
self.load_format = "runai_streamer"
if self.disable_mm_preprocessor_cache:
logger.warning(
"`--disable-mm-preprocessor-cache` is deprecated "
"and will be removed in v0.13. "
"Please use `--mm-processor-cache-gb 0` instead.", )
self.mm_processor_cache_gb = 0
elif envs.VLLM_MM_INPUT_CACHE_GIB != 4:
logger.warning(
"VLLM_MM_INPUT_CACHE_GIB` is deprecated "
"and will be removed in v0.13. "
"Please use `--mm-processor-cache-gb %d` instead.",
envs.VLLM_MM_INPUT_CACHE_GIB,
)
self.mm_processor_cache_gb = envs.VLLM_MM_INPUT_CACHE_GIB
if self.enable_multimodal_encoder_data_parallel:
logger.warning(
"--enable-multimodal-encoder-data-parallel` is deprecated "
"and will be removed in v0.13. "
"Please use `--mm-encoder-tp-mode data` instead.")
self.mm_encoder_tp_mode = "data"
return ModelConfig(
model=self.model,
hf_config_path=self.hf_config_path,
runner=self.runner,
convert=self.convert,
task=self.task,
tokenizer=self.tokenizer,
tokenizer_mode=self.tokenizer_mode,
trust_remote_code=self.trust_remote_code,
allowed_local_media_path=self.allowed_local_media_path,
dtype=self.dtype,
seed=self.seed,
revision=self.revision,
code_revision=self.code_revision,
rope_scaling=self.rope_scaling,
rope_theta=self.rope_theta,
hf_token=self.hf_token,
hf_overrides=self.hf_overrides,
tokenizer_revision=self.tokenizer_revision,
max_model_len=self.max_model_len,
quantization=self.quantization,
enforce_eager=self.enforce_eager,
max_seq_len_to_capture=self.max_seq_len_to_capture,
max_logprobs=self.max_logprobs,
logprobs_mode=self.logprobs_mode,
disable_sliding_window=self.disable_sliding_window,
disable_cascade_attn=self.disable_cascade_attn,
skip_tokenizer_init=self.skip_tokenizer_init,
enable_prompt_embeds=self.enable_prompt_embeds,
served_model_name=self.served_model_name,
limit_mm_per_prompt=self.limit_mm_per_prompt,
interleave_mm_strings=self.interleave_mm_strings,
media_io_kwargs=self.media_io_kwargs,
skip_mm_profiling=self.skip_mm_profiling,
use_async_output_proc=not self.disable_async_output_proc,
config_format=self.config_format,
mm_processor_kwargs=self.mm_processor_kwargs,
mm_processor_cache_gb=self.mm_processor_cache_gb,
mm_encoder_tp_mode=self.mm_encoder_tp_mode,
override_neuron_config=self.override_neuron_config,
override_pooler_config=self.override_pooler_config,
logits_processor_pattern=self.logits_processor_pattern,
generation_config=self.generation_config,
override_generation_config=self.override_generation_config,
enable_sleep_mode=self.enable_sleep_mode,
model_impl=self.model_impl,
override_attention_dtype=self.override_attention_dtype,
logits_processors=self.logits_processors,
)
def validate_tensorizer_args(self):
from vllm.model_executor.model_loader.tensorizer import (
TensorizerConfig)
for key in self.model_loader_extra_config:
if key in TensorizerConfig._fields:
self.model_loader_extra_config["tensorizer_config"][
key] = self.model_loader_extra_config[key]
def create_load_config(self) -> LoadConfig:
if self.quantization == "bitsandbytes":
self.load_format = "bitsandbytes"