Russell Bryant
e489ad7a21
- **Add SPDX license headers to python source files** - **Check for SPDX headers using pre-commit** commit 9d7ef44c3cfb72ca4c32e1c677d99259d10d4745 Author: Russell Bryant <rbryant@redhat.com> Date: Fri Jan 31 14:18:24 2025 -0500 Add SPDX license headers to python source files This commit adds SPDX license headers to python source files as recommended to the project by the Linux Foundation. These headers provide a concise way that is both human and machine readable for communicating license information for each source file. It helps avoid any ambiguity about the license of the code and can also be easily used by tools to help manage license compliance. The Linux Foundation runs license scans against the codebase to help ensure we are in compliance with the licenses of the code we use, including dependencies. Having these headers in place helps that tool do its job. More information can be found on the SPDX site: - https://spdx.dev/learn/handling-license-info/ Signed-off-by: Russell Bryant <rbryant@redhat.com> commit 5a1cf1cb3b80759131c73f6a9dddebccac039dea Author: Russell Bryant <rbryant@redhat.com> Date: Fri Jan 31 14:36:32 2025 -0500 Check for SPDX headers using pre-commit Signed-off-by: Russell Bryant <rbryant@redhat.com> --------- Signed-off-by: Russell Bryant <rbryant@redhat.com>
59 lines
2.0 KiB
Python
59 lines
2.0 KiB
Python
# SPDX-License-Identifier: Apache-2.0
|
|
|
|
from abc import ABC, abstractmethod
|
|
from typing import List, Optional, Set, Tuple
|
|
|
|
from vllm.model_executor.layers.sampler import SamplerOutput
|
|
from vllm.sequence import ExecuteModelRequest
|
|
from vllm.spec_decode.interfaces import SpeculativeProposer
|
|
from vllm.worker.worker_base import LoraNotSupportedWorkerBase
|
|
|
|
|
|
class ProposerWorkerBase(LoraNotSupportedWorkerBase, SpeculativeProposer):
|
|
"""Interface for proposer workers"""
|
|
|
|
@abstractmethod
|
|
def sampler_output(
|
|
self,
|
|
execute_model_req: ExecuteModelRequest,
|
|
sample_len: int,
|
|
# A set containing all sequence IDs that were assigned bonus tokens
|
|
# in their last forward pass. This set is used to backfill the KV cache
|
|
# with the key-value pairs of the penultimate token in the sequences.
|
|
# This parameter is only used by the MultiStepWorker, which relies on
|
|
# the KV cache for token generation. It is not used by workers that
|
|
# do not utilize the KV cache.
|
|
seq_ids_with_bonus_token_in_last_step: Set[int]
|
|
) -> Tuple[Optional[List[SamplerOutput]], bool]:
|
|
raise NotImplementedError
|
|
|
|
def set_include_gpu_probs_tensor(self) -> None:
|
|
"""Implementation optional"""
|
|
pass
|
|
|
|
def set_should_modify_greedy_probs_inplace(self) -> None:
|
|
"""Implementation optional"""
|
|
pass
|
|
|
|
|
|
class NonLLMProposerWorkerBase(ProposerWorkerBase, ABC):
|
|
"""Proposer worker which does not use a model with kvcache"""
|
|
|
|
def execute_model(
|
|
self,
|
|
execute_model_req: Optional[ExecuteModelRequest] = None
|
|
) -> List[SamplerOutput]:
|
|
"""get_spec_proposals is used to get the proposals"""
|
|
return []
|
|
|
|
def determine_num_available_blocks(self) -> Tuple[int, int]:
|
|
"""This is never called on the proposer, only the target model"""
|
|
raise NotImplementedError
|
|
|
|
def initialize_cache(self, num_gpu_blocks: int,
|
|
num_cpu_blocks: int) -> None:
|
|
pass
|
|
|
|
def get_cache_block_size_bytes(self) -> int:
|
|
return 0
|