Skip to content

vllm_mlx.tool_parsers.granite_tool_parser

Granite tool call parser for vllm-mlx.

View the complete module source at #L1-L147.

API details

Each callable below includes its exact signature, type annotations, inputs, defaults, return contract, documented exceptions, implementation source, and parsed docstring sections when the source provides them.

vllm_mlx.tool_parsers.granite_tool_parser

Granite tool call parser for vllm-mlx.

Handles IBM Granite models' tool calling format: - <|tool_call|> or followed by JSON array - [{"name": "func", "arguments": {...}}]

vllm_mlx.tool_parsers.granite_tool_parser.GraniteToolParser 

GraniteToolParser(tokenizer: PreTrainedTokenizerBase | None = None)

Bases: ToolParser

Tool call parser for IBM Granite models.

Supports Granite's tool call format: <|tool_call|>[{"name": "get_weather", "arguments": {"city": "Paris"}}]

Or Granite 3.1: [{"name": "get_weather", "arguments": {"city": "Paris"}}]

Used when --enable-auto-tool-choice --tool-call-parser granite are set.

Source code in vllm_mlx/tool_parsers/abstract_tool_parser.py 
103
104
105
106
107
108
109
110
111
112
113
def __init__(self, tokenizer: PreTrainedTokenizerBase | None = None):
    """
    Initialize the tool parser.

    Args:
        tokenizer: The tokenizer for the model (optional, some parsers need it)
    """
    self.model_tokenizer = tokenizer
    # State for streaming parsing
    self.current_tool_id: int = -1
    self.prev_tool_call_arr: list[dict] = []

vllm_mlx.tool_parsers.granite_tool_parser.GraniteToolParser.SUPPORTS_NATIVE_TOOL_FORMAT class-attribute instance-attribute 

SUPPORTS_NATIVE_TOOL_FORMAT = True

vllm_mlx.tool_parsers.granite_tool_parser.GraniteToolParser.BOT_TOKEN class-attribute instance-attribute 

BOT_TOKEN = '<|tool_call|>'

vllm_mlx.tool_parsers.granite_tool_parser.GraniteToolParser.BOT_STRING class-attribute instance-attribute 

BOT_STRING = '<tool_call>'

vllm_mlx.tool_parsers.granite_tool_parser.GraniteToolParser.extract_tool_calls 

extract_tool_calls(model_output: str, request: dict[str, Any] | None = None) -> ExtractedToolCallInformation

Extract tool calls from Granite model output.

Source code in vllm_mlx/tool_parsers/granite_tool_parser.py 
 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
def extract_tool_calls(
    self, model_output: str, request: dict[str, Any] | None = None
) -> ExtractedToolCallInformation:
    """
    Extract tool calls from Granite model output.
    """
    stripped = model_output.strip()

    # Remove tool call markers
    if stripped.startswith(self.BOT_TOKEN):
        stripped = stripped[len(self.BOT_TOKEN) :].lstrip()
    elif stripped.startswith(self.BOT_STRING):
        stripped = stripped[len(self.BOT_STRING) :].lstrip()

    # Check if it starts with JSON array
    if not stripped or stripped[0] != "[":
        return ExtractedToolCallInformation(
            tools_called=False, tool_calls=[], content=model_output
        )

    try:
        raw_calls = json.loads(stripped)
        if not isinstance(raw_calls, list):
            return ExtractedToolCallInformation(
                tools_called=False, tool_calls=[], content=model_output
            )

        tool_calls = []
        for call in raw_calls:
            if isinstance(call, dict):
                # Granite uses "name" or "type" for function name
                func_name = call.get("name") or call.get("type")
                if func_name:
                    args = call.get("arguments", {})
                    tool_calls.append(
                        {
                            "id": generate_tool_id(),
                            "name": func_name,
                            "arguments": (
                                json.dumps(args, ensure_ascii=False)
                                if isinstance(args, dict)
                                else str(args)
                            ),
                        }
                    )

        if tool_calls:
            return ExtractedToolCallInformation(
                tools_called=True,
                tool_calls=tool_calls,
                content=None,
            )

    except json.JSONDecodeError:
        pass

    return ExtractedToolCallInformation(
        tools_called=False, tool_calls=[], content=model_output
    )

vllm_mlx.tool_parsers.granite_tool_parser.GraniteToolParser.extract_tool_calls_streaming 

extract_tool_calls_streaming(previous_text: str, current_text: str, delta_text: str, previous_token_ids: Sequence[int] | None = None, current_token_ids: Sequence[int] | None = None, delta_token_ids: Sequence[int] | None = None, request: dict[str, Any] | None = None) -> dict[str, Any] | None

Extract tool calls from streaming Granite model output.

Source code in vllm_mlx/tool_parsers/granite_tool_parser.py 
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
def extract_tool_calls_streaming(
    self,
    previous_text: str,
    current_text: str,
    delta_text: str,
    previous_token_ids: Sequence[int] | None = None,
    current_token_ids: Sequence[int] | None = None,
    delta_token_ids: Sequence[int] | None = None,
    request: dict[str, Any] | None = None,
) -> dict[str, Any] | None:
    """
    Extract tool calls from streaming Granite model output.
    """
    stripped = current_text.strip()
    has_marker = stripped.startswith(self.BOT_TOKEN) or stripped.startswith(
        self.BOT_STRING
    )

    if not has_marker:
        return {"content": delta_text}

    # Try to parse when we have a complete JSON array
    if "]" in delta_text:
        result = self.extract_tool_calls(current_text)
        if result.tools_called:
            return {
                "tool_calls": [
                    {
                        "index": i,
                        "id": tc["id"],
                        "type": "function",
                        "function": {
                            "name": tc["name"],
                            "arguments": tc["arguments"],
                        },
                    }
                    for i, tc in enumerate(result.tool_calls)
                ]
            }

    return None

vllm_mlx.tool_parsers.granite_tool_parser.generate_tool_id 

generate_tool_id() -> str

Generate a unique tool call ID.

Source code in vllm_mlx/tool_parsers/granite_tool_parser.py 
22
23
24
def generate_tool_id() -> str:
    """Generate a unique tool call ID."""
    return f"call_{uuid.uuid4().hex[:8]}"

Complete contract reference

Expand any definition for its exact inputs, annotations, defaults, return contract, directly raised exceptions, source-grounded behavior, and immutable line link. This section includes private and nested definitions that ordinary API generators omit.

vllm_mlx.tool_parsers.granite_tool_parser.generate_tool_id · function 
vllm_mlx.tool_parsers.granite_tool_parser.generate_tool_id() -> str

Generate a unique tool call ID.

Parameters

This callable has no explicit inputs.

Returns

  • Type: str
  • Direct return expressions: f'call_{uuid.uuid4().hex[:8]}'

Exceptions and behavior

Function generate_tool_id calls uuid.uuid4; returns f'call_{uuid.uuid4().hex[:8]}'. No direct raise statement appears in this definition.

View source #L22-L24.

vllm_mlx.tool_parsers.granite_tool_parser.GraniteToolParser · class 
vllm_mlx.tool_parsers.granite_tool_parser.GraniteToolParser()

Tool call parser for IBM Granite models.

Parameters

This callable has no explicit inputs.

Returns

  • Constructs: vllm_mlx.tool_parsers.granite_tool_parser.GraniteToolParser

Exceptions and behavior

Class GraniteToolParser derives from ToolParser and declares 2 direct member(s). No direct raise statement appears in this definition.

View source #L28-L147.

vllm_mlx.tool_parsers.granite_tool_parser.GraniteToolParser.extract_tool_calls · method 
vllm_mlx.tool_parsers.granite_tool_parser.GraniteToolParser.extract_tool_calls(model_output: str, request: dict[str, Any] | None = None) -> ExtractedToolCallInformation

Extract tool calls from Granite model output.

Parameters

Name Type Required Default Description
model_output str yes none Required positional or keyword input.
request dict[str, Any] \| None no None Optional positional or keyword input; defaults to None.

Returns

  • Type: ExtractedToolCallInformation
  • Direct return expressions: ExtractedToolCallInformation(tools_called=False, tool_calls=[], content=model_output); ExtractedToolCallInformation(tools_called=True, tool_calls=tool_calls, content=None)

Exceptions and behavior

Method GraniteToolParser.extract_tool_calls calls model_output.strip, stripped.startswith, stripped[len(self.BOT_TOKEN):].lstrip, len; has 2 explicit return paths. No direct raise statement appears in this definition.

View source #L47-L105.

vllm_mlx.tool_parsers.granite_tool_parser.GraniteToolParser.extract_tool_calls_streaming · method 
vllm_mlx.tool_parsers.granite_tool_parser.GraniteToolParser.extract_tool_calls_streaming(previous_text: str, current_text: str, delta_text: str, previous_token_ids: Sequence[int] | None = None, current_token_ids: Sequence[int] | None = None, delta_token_ids: Sequence[int] | None = None, request: dict[str, Any] | None = None) -> dict[str, Any] | None

Extract tool calls from streaming Granite model output.

Parameters

Name Type Required Default Description
previous_text str yes none Required positional or keyword input.
current_text str yes none Required positional or keyword input.
delta_text str yes none Required positional or keyword input.
previous_token_ids Sequence[int] \| None no None Optional positional or keyword input; defaults to None.
current_token_ids Sequence[int] \| None no None Optional positional or keyword input; defaults to None.
delta_token_ids Sequence[int] \| None no None Optional positional or keyword input; defaults to None.
request dict[str, Any] \| None no None Optional positional or keyword input; defaults to None.

Returns

  • Type: dict[str, Any] | None
  • Direct return expressions: {'content': delta_text}; {'tool_calls': [{'index': i, 'id': tc['id'], 'type': 'function', 'function': {'name': tc['name'], 'arguments': tc['argu…; None

Exceptions and behavior

Method GraniteToolParser.extract_tool_calls_streaming calls current_text.strip, stripped.startswith, self.extract_tool_calls, enumerate; has 3 explicit return paths. No direct raise statement appears in this definition.

View source #L107-L147.

Complete symbol map

This map also includes private definitions and nested helpers. The signature column exposes every explicit input even when an internal helper has no dedicated parameter prose.

Symbol Kind Signature and inputs What it does Source
generate_tool_id function generate_tool_id() -> str Generate a unique tool call ID. #L22-L24
GraniteToolParser class GraniteToolParser() Tool call parser for IBM Granite models. #L28-L147
GraniteToolParser.extract_tool_calls method GraniteToolParser.extract_tool_calls(model_output: str, request: dict[str, Any] \| None = None) -> ExtractedToolCallInformation Extract tool calls from Granite model output. #L47-L105
GraniteToolParser.extract_tool_calls_streaming method GraniteToolParser.extract_tool_calls_streaming(previous_text: str, current_text: str, delta_text: str, previous_token_ids: Sequence[int] \| None = None, current_token_ids: Sequence[int] \| None = None, delta_token_ids: Sequence[int] \| None = None, request: dict[str, Any] \| None = None) -> dict[str, Any] \| None Extract tool calls from streaming Granite model output. #L107-L147