MCPcopy
hub / github.com/NVIDIA/TensorRT-LLM / GuidedDecodingConfig

Class GuidedDecodingConfig

tensorrt_llm/llmapi/llm_args.py:167–184  ·  view source on GitHub ↗

Source from the content-addressed store, hash-verified

165
166
167class GuidedDecodingConfig(StrictBaseModel):
168
169 class GuidedDecodingBackend(Enum):
170 XGRAMMAR = 0
171 LLGUIDANCE = 1
172
173 backend: GuidedDecodingBackend = Field(
174 default=GuidedDecodingBackend.XGRAMMAR,
175 description="The backend for guided decoding config.")
176 encoded_vocab: Optional[List[str]] = Field(
177 default=None,
178 description="The encoded vocab for guided decoding config.")
179 tokenizer_str: Optional[str] = Field(
180 default=None,
181 description="The tokenizer string for guided decoding config.")
182 stop_token_ids: Optional[List[int]] = Field(
183 default=None,
184 description="The stop token ids for guided decoding config.")
185
186
187class BaseSparseAttentionConfig(StrictBaseModel):

Callers 1

Calls 1

FieldFunction · 0.85

Tested by

no test coverage detected