biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

Add full API docs and improve the UX of navigating them (#17485)

Browse Source 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-05-04 03:42:43 +01:00
committed by GitHub
parent 46fae69cf0
commit d6484ef3c3
101 changed files with 872 additions and 980 deletions
Download Patch File Download Diff File Expand all files Collapse all files

62
vllm/entrypoints/llm.py
View File

@@ -115,7 +115,7 @@ class LLM:
to eager mode. Additionally for encoder-decoder models, if the
sequence length of the encoder input is larger than this, we fall
back to the eager mode.
disable_custom_all_reduce: See :class:`~vllm.config.ParallelConfig`
disable_custom_all_reduce: See {class}`~vllm.config.ParallelConfig`
disable_async_output_proc: Disable async output processing.
This may result in lower performance.
hf_token: The token to use as HTTP bearer authorization for remote files
@@ -127,12 +127,13 @@ class LLM:
compilation_config: Either an integer or a dictionary. If it is an
integer, it is used as the level of compilation optimization. If it
is a dictionary, it can specify the full compilation configuration.
**kwargs: Arguments for :class:`~vllm.EngineArgs`. (See
:ref:`engine-args`)
**kwargs: Arguments for {class}`~vllm.EngineArgs`. (See
{ref}`engine-args`)
Note:
This class is intended to be used for offline inference. For online
serving, use the :class:`~vllm.AsyncLLMEngine` class instead.
:::{note}
This class is intended to be used for offline inference. For online
serving, use the {class}`~vllm.AsyncLLMEngine` class instead.
:::
"""
DEPRECATE_LEGACY: ClassVar[bool] = True
@@ -141,7 +142,7 @@ class LLM:
DEPRECATE_INIT_POSARGS: ClassVar[bool] = True
"""
A flag to toggle whether to deprecate positional arguments in
:meth:`LLM.__init__`.
{meth}`LLM.__init__`.
"""
@classmethod
@@ -398,7 +399,7 @@ class LLM:
Args:
prompts: The prompts to the LLM. You may pass a sequence of prompts
for batch inference. See :class:`~vllm.inputs.PromptType`
for batch inference. See {class}`~vllm.inputs.PromptType`
for more details about the format of each prompts.
sampling_params: The sampling parameters for text generation. If
None, we use the default sampling parameters.
@@ -413,13 +414,14 @@ class LLM:
Only applicable when priority scheduling policy is enabled.
Returns:
A list of ``RequestOutput`` objects containing the
A list of `RequestOutput` objects containing the
generated completions in the same order as the input prompts.
Note:
Using ``prompts`` and ``prompt_token_ids`` as keyword parameters is
considered legacy and may be deprecated in the future. You should
instead pass them via the ``inputs`` parameter.
:::{note}
Using `prompts` and `prompt_token_ids` as keyword parameters is
considered legacy and may be deprecated in the future. You should
instead pass them via the `inputs` parameter.
:::
"""
runner_type = self.llm_engine.model_config.runner_type
if runner_type not in ["generate", "transcription"]:
@@ -488,16 +490,17 @@ class LLM:
`self` argument, in addition to the arguments passed in `args`
and `kwargs`. The `self` argument will be the worker object.
timeout: Maximum time in seconds to wait for execution. Raises a
:exc:`TimeoutError` on timeout. `None` means wait indefinitely.
{exc}`TimeoutError` on timeout. `None` means wait indefinitely.
args: Positional arguments to pass to the worker method.
kwargs: Keyword arguments to pass to the worker method.
Returns:
A list containing the results from each worker.
Note:
It is recommended to use this API to only pass control messages,
and set up data-plane communication to pass data.
:::{note}
It is recommended to use this API to only pass control messages,
and set up data-plane communication to pass data.
:::
"""
return self.llm_engine.collective_rpc(method, timeout, args, kwargs)
@@ -664,7 +667,7 @@ class LLM:
Generate responses for a chat conversation.
The chat conversation is converted into a text prompt using the
tokenizer and calls the :meth:`generate` method to generate the
tokenizer and calls the {meth}`generate` method to generate the
responses.
Multi-modal inputs can be passed in the same way you would pass them
@@ -903,7 +906,7 @@ class LLM:
Args:
prompts: The prompts to the LLM. You may pass a sequence of prompts
for batch inference. See :class:`~vllm.inputs.PromptType`
for batch inference. See {class}`~vllm.inputs.PromptType`
for more details about the format of each prompts.
pooling_params: The pooling parameters for pooling. If None, we
use the default pooling parameters.
@@ -913,13 +916,14 @@ class LLM:
generation, if any.
Returns:
A list of ``PoolingRequestOutput`` objects containing the
A list of `PoolingRequestOutput` objects containing the
pooled hidden states in the same order as the input prompts.
Note:
Using ``prompts`` and ``prompt_token_ids`` as keyword parameters is
considered legacy and may be deprecated in the future. You should
instead pass them via the ``inputs`` parameter.
:::{note}
Using `prompts` and `prompt_token_ids` as keyword parameters is
considered legacy and may be deprecated in the future. You should
instead pass them via the `inputs` parameter.
:::
"""
runner_type = self.llm_engine.model_config.runner_type
if runner_type != "pooling":
@@ -992,7 +996,7 @@ class LLM:
Args:
prompts: The prompts to the LLM. You may pass a sequence of prompts
for batch inference. See :class:`~vllm.inputs.PromptType`
for batch inference. See {class}`~vllm.inputs.PromptType`
for more details about the format of each prompts.
pooling_params: The pooling parameters for pooling. If None, we
use the default pooling parameters.
@@ -1036,7 +1040,7 @@ class LLM:
Args:
prompts: The prompts to the LLM. You may pass a sequence of prompts
for batch inference. See :class:`~vllm.inputs.PromptType`
for batch inference. See {class}`~vllm.inputs.PromptType`
for more details about the format of each prompts.
use_tqdm: Whether to use tqdm to display the progress bar.
lora_request: LoRA request to use for generation, if any.
@@ -1168,7 +1172,7 @@ class LLM:
text_1: can be a single prompt or a list of prompts, in which
case it has to have the same length as the ``text_2`` list
text_2: The texts to pair with the query to form the input
to the LLM. See :class:`~vllm.inputs.PromptType` for
to the LLM. See {class}`~vllm.inputs.PromptType` for
more details about the format of each prompts.
use_tqdm: Whether to use tqdm to display the progress bar.
lora_request: LoRA request to use for generation, if any.
@@ -1277,7 +1281,7 @@ class LLM:
def wake_up(self, tags: Optional[list[str]] = None):
"""
Wake up the engine from sleep mode. See the :meth:`sleep` method
Wake up the engine from sleep mode. See the {meth}`sleep` method
for more details.
Args:

19
vllm/entrypoints/openai/protocol.py
View File

@@ -5,7 +5,6 @@
import json
import re
import time
from argparse import Namespace
from typing import Annotated, Any, ClassVar, Literal, Optional, Union
import torch
@@ -25,23 +24,7 @@ from vllm.utils import random_uuid, resolve_obj_by_qualname
logger = init_logger(__name__)
# torch is mocked during docs generation,
# so we have to provide the values as literals
_MOCK_LONG_INFO = Namespace(min=-9223372036854775808, max=9223372036854775807)
_LONG_INFO: Union["torch.iinfo", Namespace]
try:
from sphinx.ext.autodoc.mock import _MockModule
if isinstance(torch, _MockModule):
_LONG_INFO = _MOCK_LONG_INFO
else:
_LONG_INFO = torch.iinfo(torch.long)
except ModuleNotFoundError:
_LONG_INFO = torch.iinfo(torch.long)
assert _LONG_INFO.min == _MOCK_LONG_INFO.min
assert _LONG_INFO.max == _MOCK_LONG_INFO.max
_LONG_INFO = torch.iinfo(torch.long)
class OpenAIBaseModel(BaseModel):

4
vllm/entrypoints/openai/serving_engine.py
View File

@@ -275,7 +275,7 @@ class OpenAIServing:
add_special_tokens: bool = True,
) -> TextTokensPrompt:
"""
A simpler implementation of :meth:`_tokenize_prompt_input_or_inputs`
A simpler implementation of {meth}`_tokenize_prompt_input_or_inputs`
that assumes single input.
"""
return next(
@@ -296,7 +296,7 @@ class OpenAIServing:
add_special_tokens: bool = True,
) -> Iterator[TextTokensPrompt]:
"""
A simpler implementation of :meth:`_tokenize_prompt_input_or_inputs`
A simpler implementation of {meth}`_tokenize_prompt_input_or_inputs`
that assumes multiple inputs.
"""
for text in prompt_inputs: