[Platform] Deprecate seed_everything (#31659)
Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
This commit is contained in:
@@ -2,10 +2,8 @@
|
||||
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
||||
|
||||
from vllm.model_executor.parameter import BasevLLMParameter, PackedvLLMParameter
|
||||
from vllm.model_executor.utils import set_random_seed
|
||||
|
||||
__all__ = [
|
||||
"set_random_seed",
|
||||
"BasevLLMParameter",
|
||||
"PackedvLLMParameter",
|
||||
]
|
||||
|
||||
@@ -10,12 +10,6 @@ import torch
|
||||
from vllm.utils.torch_utils import is_torch_equal_or_newer
|
||||
|
||||
|
||||
def set_random_seed(seed: int | None) -> None:
|
||||
from vllm.platforms import current_platform
|
||||
|
||||
current_platform.seed_everything(seed)
|
||||
|
||||
|
||||
def set_weight_attrs(
|
||||
weight: torch.Tensor,
|
||||
weight_attrs: dict[str, Any] | None,
|
||||
|
||||
@@ -372,6 +372,10 @@ class Platform:
|
||||
|
||||
Loosely based on: https://github.com/Lightning-AI/pytorch-lightning/blob/2.4.0/src/lightning/fabric/utilities/seed.py#L20
|
||||
"""
|
||||
logger.info_once(
|
||||
"`seed_everything` is deprecated. It will be removed in v0.14.0 or later. "
|
||||
"Please use `vllm.utils.torch_utils.set_random_seed` instead."
|
||||
)
|
||||
if seed is not None:
|
||||
random.seed(seed)
|
||||
np.random.seed(seed)
|
||||
|
||||
@@ -3,6 +3,7 @@
|
||||
import contextlib
|
||||
import importlib.metadata
|
||||
import os
|
||||
import random
|
||||
import threading
|
||||
from collections.abc import Callable, Collection
|
||||
from functools import lru_cache
|
||||
@@ -278,6 +279,13 @@ def kv_cache_dtype_str_to_dtype(
|
||||
return STR_DTYPE_TO_TORCH_DTYPE[kv_cache_dtype]
|
||||
|
||||
|
||||
def set_random_seed(seed: int | None) -> None:
|
||||
if seed is not None:
|
||||
random.seed(seed)
|
||||
np.random.seed(seed)
|
||||
torch.manual_seed(seed)
|
||||
|
||||
|
||||
def create_kv_caches_with_random_flash(
|
||||
num_blocks: int,
|
||||
block_size: int,
|
||||
@@ -290,9 +298,7 @@ def create_kv_caches_with_random_flash(
|
||||
device: str | None = "cuda",
|
||||
cache_layout: str | None = "NHD",
|
||||
) -> tuple[list[torch.Tensor], list[torch.Tensor]]:
|
||||
from vllm.platforms import current_platform
|
||||
|
||||
current_platform.seed_everything(seed)
|
||||
set_random_seed(seed)
|
||||
|
||||
dtype = get_kv_cache_torch_dtype(cache_dtype, model_dtype)
|
||||
generic_kv_cache_shape = (num_blocks, 2, block_size, num_heads, head_size)
|
||||
@@ -335,9 +341,8 @@ def create_kv_caches_with_random(
|
||||
raise ValueError(
|
||||
f"Does not support key cache of type fp8 with head_size {head_size}"
|
||||
)
|
||||
from vllm.platforms import current_platform
|
||||
|
||||
current_platform.seed_everything(seed)
|
||||
set_random_seed(seed)
|
||||
|
||||
dtype = get_kv_cache_torch_dtype(cache_dtype, model_dtype)
|
||||
|
||||
|
||||
@@ -10,10 +10,10 @@ import torch
|
||||
from vllm import envs
|
||||
from vllm.config import VllmConfig
|
||||
from vllm.logger import init_logger
|
||||
from vllm.model_executor.utils import set_random_seed
|
||||
from vllm.platforms import CpuArchEnum, current_platform
|
||||
from vllm.platforms.cpu import CpuPlatform, LogicalCPUInfo
|
||||
from vllm.profiler.wrapper import TorchProfilerWrapper
|
||||
from vllm.utils.torch_utils import set_random_seed
|
||||
from vllm.v1.worker.cpu_model_runner import CPUModelRunner
|
||||
from vllm.v1.worker.gpu_worker import Worker, init_worker_distributed_environment
|
||||
|
||||
|
||||
@@ -34,7 +34,6 @@ from vllm.distributed.parallel_state import (
|
||||
)
|
||||
from vllm.logger import init_logger
|
||||
from vllm.lora.request import LoRARequest
|
||||
from vllm.model_executor import set_random_seed
|
||||
from vllm.model_executor.models.interfaces import is_mixture_of_experts
|
||||
from vllm.model_executor.warmup.kernel_warmup import kernel_warmup
|
||||
from vllm.platforms import current_platform
|
||||
@@ -43,6 +42,7 @@ from vllm.sequence import IntermediateTensors
|
||||
from vllm.tasks import SupportedTask
|
||||
from vllm.utils.mem_constants import GiB_bytes
|
||||
from vllm.utils.mem_utils import MemorySnapshot, memory_profiling
|
||||
from vllm.utils.torch_utils import set_random_seed
|
||||
from vllm.v1.core.sched.output import GrammarOutput, SchedulerOutput
|
||||
from vllm.v1.engine import ReconfigureDistributedRequest, ReconfigureRankType
|
||||
from vllm.v1.kv_cache_interface import KVCacheConfig, KVCacheSpec
|
||||
|
||||
@@ -20,12 +20,11 @@ from vllm.distributed.kv_transfer import (
|
||||
)
|
||||
from vllm.logger import init_logger
|
||||
from vllm.lora.request import LoRARequest
|
||||
from vllm.model_executor import set_random_seed
|
||||
from vllm.platforms import current_platform
|
||||
from vllm.platforms.tpu import USE_TPU_INFERENCE
|
||||
from vllm.tasks import SupportedTask
|
||||
from vllm.utils.math_utils import cdiv
|
||||
from vllm.utils.torch_utils import STR_DTYPE_TO_TORCH_DTYPE
|
||||
from vllm.utils.torch_utils import STR_DTYPE_TO_TORCH_DTYPE, set_random_seed
|
||||
from vllm.v1.core.sched.output import GrammarOutput, SchedulerOutput
|
||||
from vllm.v1.kv_cache_interface import AttentionSpec, KVCacheConfig, KVCacheSpec
|
||||
from vllm.v1.outputs import ModelRunnerOutput
|
||||
|
||||
@@ -9,9 +9,9 @@ import torch.distributed
|
||||
from vllm.config import VllmConfig
|
||||
from vllm.distributed import get_world_group
|
||||
from vllm.logger import init_logger
|
||||
from vllm.model_executor import set_random_seed
|
||||
from vllm.platforms import current_platform
|
||||
from vllm.profiler.wrapper import TorchProfilerWrapper
|
||||
from vllm.utils.torch_utils import set_random_seed
|
||||
from vllm.v1.worker.gpu_worker import Worker, init_worker_distributed_environment
|
||||
from vllm.v1.worker.xpu_model_runner import XPUModelRunner
|
||||
|
||||
|
||||
Reference in New Issue
Block a user