[Doc] Improve documentation for multimodal CLI args (vllm-project#16960)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Signed-off-by: Agata Dobrzyniewicz <adobrzyniewicz@habana.ai>

Author: DarkLight1337

vllm/config.py

@@ -54,13 +54,15 @@
     from vllm.transformers_utils.tokenizer_group.base_tokenizer_group import (
         BaseTokenizerGroup)
 
-    Config = TypeVar("Config", bound=DataclassInstance)
+    ConfigType = type[DataclassInstance]
 else:
     QuantizationConfig = None
-    Config = TypeVar("Config")
+    ConfigType = type
 
 logger = init_logger(__name__)
 
+ConfigT = TypeVar("ConfigT", bound=ConfigType)
+
 # This value is chosen to have a balance between ITL and TTFT. Note it is
 # not optimized for throughput.
 _DEFAULT_MAX_NUM_BATCHED_TOKENS = 2048
@@ -162,7 +164,7 @@ def pairwise(iterable):
     return out
 
 
-def config(cls: type[Config]) -> type[Config]:
+def config(cls: ConfigT) -> ConfigT:
     """
     A decorator that ensures all fields in a dataclass have default values
     and that each field has a docstring.
@@ -181,7 +183,7 @@ def config(cls: type[Config]) -> type[Config]:
     return cls
 
 
-def get_field(cls: type[Config], name: str) -> Field:
+def get_field(cls: ConfigType, name: str) -> Field:
     """Get the default factory field of a dataclass by name. Used for getting
     default factory fields in `EngineArgs`."""
     if not is_dataclass(cls):
@@ -2749,6 +2751,9 @@ class MultiModalConfig:
     The maximum number of input items allowed per prompt for each modality.
     This should be a JSON string that will be parsed into a dictionary.
     Defaults to 1 (V0) or 999 (V1) for each modality.
+
+    For example, to allow up to 16 images and 2 videos per prompt:
+    ``{"images": 16, "videos": 2}``
     """
 
     def compute_hash(self) -> str:

vllm/engine/arg_utils.py

@@ -17,7 +17,7 @@
 import vllm.envs as envs
 from vllm import version
 from vllm.config import (BlockSize, CacheConfig, CacheDType, CompilationConfig,
-                         Config, ConfigFormat, DecodingConfig, Device,
+                         ConfigFormat, ConfigType, DecodingConfig, Device,
                          DeviceConfig, DistributedExecutorBackend, HfOverrides,
                          KVTransferConfig, LoadConfig, LoadFormat, LoRAConfig,
                          ModelConfig, ModelImpl, MultiModalConfig,
@@ -304,7 +304,7 @@ def is_custom_type(cls: TypeHint) -> bool:
             """Check if the class is a custom type."""
             return cls.__module__ != "builtins"
 
-        def get_kwargs(cls: type[Config]) -> dict[str, Any]:
+        def get_kwargs(cls: ConfigType) -> dict[str, Any]:
             cls_docs = get_attr_docs(cls)
             kwargs = {}
             for field in fields(cls):
@@ -678,13 +678,15 @@ def get_kwargs(cls: type[Config]) -> dict[str, Any]:
             '--mm-processor-kwargs',
             default=None,
             type=json.loads,
-            help=('Overrides for the multimodal input mapping/processing, '
-                  'e.g., image processor. For example: ``{"num_crops": 4}``.'))
+            help=('Overrides for the multi-modal processor obtained from '
+                  '``AutoProcessor.from_pretrained``. The available overrides '
+                  'depend on the model that is being run.'
+                  'For example, for Phi-3-Vision: ``{"num_crops": 4}``.'))
         parser.add_argument(
             '--disable-mm-preprocessor-cache',
             action='store_true',
-            help='If true, then disables caching of the multi-modal '
-            'preprocessor/mapper. (not recommended)')
+            help='If True, disable caching of the processed multi-modal '
+            'inputs.')
 
         # LoRA related configs
         parser.add_argument('--enable-lora',