iHore/vLLM
8
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[Deprecation] Mark env vars covered by --moe-backend / --linear-backend (#43148)

Browse Source 
Signed-off-by: mgoin <mgoin64@gmail.com>
Signed-off-by: Michael Goin <mgoin64@gmail.com>
This commit is contained in:
Michael Goin
2026-05-21 12:51:12 -07:00
committed by GitHub
parent 1c78f76c29
commit 9b54e50e2c
2 changed files with 113 additions and 54 deletions
Download Patch File Download Diff File Expand all files Collapse all files
vllm/envs.py
+112 -33
View File
@@ -8,6 +8,7 @@ import os
import sys
import tempfile
import uuid
import warnings
from collections.abc import Callable
from typing import TYPE_CHECKING, Any, Literal
@@ -342,6 +343,27 @@ def use_mega_aot_artifact():
return os.environ.get("VLLM_USE_MEGA_AOT_ARTIFACT", default_value) == "1"
def deprecated_env(
env_name: str,
removal_version: str,
replacement: str,
getter: Callable[[], Any],
) -> Callable[[], Any]:
"""Wrap an env-var getter to emit a FutureWarning when the var is set."""
def _read() -> Any:
if env_name in os.environ:
warnings.warn(
f"{env_name} is deprecated and will be removed in "
f"{removal_version}. {replacement}",
FutureWarning,
stacklevel=2,
)
return getter()
return _read
def env_with_choices(
env_name: str,
default: str | None,
@@ -1294,8 +1316,13 @@ environment_variables: dict[str, Callable[[], Any]] = {
os.environ.get("VLLM_MARLIN_USE_ATOMIC_ADD", "0") == "1"
),
# Whether to use marlin kernel in mxfp4 quantization method
"VLLM_MXFP4_USE_MARLIN": lambda: maybe_convert_bool(
os.environ.get("VLLM_MXFP4_USE_MARLIN", None)
# Deprecated: use --moe-backend marlin (MoE) or --linear-backend marlin
# (linear) instead.
"VLLM_MXFP4_USE_MARLIN": deprecated_env(
"VLLM_MXFP4_USE_MARLIN",
"v0.23",
"Use --moe-backend marlin or --linear-backend marlin.",
lambda: maybe_convert_bool(os.environ.get("VLLM_MXFP4_USE_MARLIN", None)),
),
# The activation dtype for marlin kernel
"VLLM_MARLIN_INPUT_DTYPE": env_with_choices(
@@ -1390,16 +1417,29 @@ environment_variables: dict[str, Callable[[], Any]] = {
int(os.getenv("VLLM_BLOCKSCALE_FP8_GEMM_FLASHINFER", "1"))
),
# Allow use of FlashInfer BF16 MoE kernels for fused moe ops.
"VLLM_USE_FLASHINFER_MOE_FP16": lambda: bool(
int(os.getenv("VLLM_USE_FLASHINFER_MOE_FP16", "0"))
# Deprecated: use --moe-backend to select a kernel explicitly.
"VLLM_USE_FLASHINFER_MOE_FP16": deprecated_env(
"VLLM_USE_FLASHINFER_MOE_FP16",
"v0.23",
"Use --moe-backend (e.g. flashinfer_trtllm, flashinfer_cutlass).",
lambda: bool(int(os.getenv("VLLM_USE_FLASHINFER_MOE_FP16", "0"))),
),
# Allow use of FlashInfer FP8 MoE kernels for fused moe ops.
"VLLM_USE_FLASHINFER_MOE_FP8": lambda: bool(
int(os.getenv("VLLM_USE_FLASHINFER_MOE_FP8", "0"))
# Deprecated: use --moe-backend to select a kernel explicitly.
"VLLM_USE_FLASHINFER_MOE_FP8": deprecated_env(
"VLLM_USE_FLASHINFER_MOE_FP8",
"v0.23",
"Use --moe-backend (e.g. flashinfer_trtllm, flashinfer_cutlass).",
lambda: bool(int(os.getenv("VLLM_USE_FLASHINFER_MOE_FP8", "0"))),
),
# Allow use of FlashInfer NVFP4 MoE kernels for fused moe ops.
"VLLM_USE_FLASHINFER_MOE_FP4": lambda: bool(
int(os.getenv("VLLM_USE_FLASHINFER_MOE_FP4", "0"))
# Deprecated: use --moe-backend to select a kernel explicitly.
"VLLM_USE_FLASHINFER_MOE_FP4": deprecated_env(
"VLLM_USE_FLASHINFER_MOE_FP4",
"v0.23",
"Use --moe-backend (e.g. flashinfer_trtllm, flashinfer_cutlass, "
"flashinfer_cutedsl).",
lambda: bool(int(os.getenv("VLLM_USE_FLASHINFER_MOE_FP4", "0"))),
),
# Allow use of FlashInfer MxInt4 MoE kernels for fused moe ops.
"VLLM_USE_FLASHINFER_MOE_INT4": lambda: bool(
@@ -1407,20 +1447,36 @@ environment_variables: dict[str, Callable[[], Any]] = {
),
# If set to 1, use the FlashInfer
# MXFP8 (activation) x MXFP4 (weight) MoE backend.
"VLLM_USE_FLASHINFER_MOE_MXFP4_MXFP8": lambda: bool(
int(os.getenv("VLLM_USE_FLASHINFER_MOE_MXFP4_MXFP8", "0"))
# Deprecated: use --moe-backend flashinfer_trtllm combined with
# --quantization_config.moe.activation mxfp8.
"VLLM_USE_FLASHINFER_MOE_MXFP4_MXFP8": deprecated_env(
"VLLM_USE_FLASHINFER_MOE_MXFP4_MXFP8",
"v0.23",
"Use --moe-backend flashinfer_trtllm with "
"--quantization_config.moe.activation mxfp8.",
lambda: bool(int(os.getenv("VLLM_USE_FLASHINFER_MOE_MXFP4_MXFP8", "0"))),
),
# If set to 1, use the FlashInfer CUTLASS backend for
# MXFP8 (activation) x MXFP4 (weight) MoE.
# This is separate from the TRTLLMGEN path controlled by
# VLLM_USE_FLASHINFER_MOE_MXFP4_MXFP8.
"VLLM_USE_FLASHINFER_MOE_MXFP4_MXFP8_CUTLASS": lambda: bool(
int(os.getenv("VLLM_USE_FLASHINFER_MOE_MXFP4_MXFP8_CUTLASS", "0"))
# Deprecated: use --moe-backend flashinfer_cutlass combined with
# --quantization_config.moe.activation mxfp8.
"VLLM_USE_FLASHINFER_MOE_MXFP4_MXFP8_CUTLASS": deprecated_env(
"VLLM_USE_FLASHINFER_MOE_MXFP4_MXFP8_CUTLASS",
"v0.23",
"Use --moe-backend flashinfer_cutlass with "
"--quantization_config.moe.activation mxfp8.",
lambda: bool(
int(os.getenv("VLLM_USE_FLASHINFER_MOE_MXFP4_MXFP8_CUTLASS", "0"))
),
),
# If set to 1, use the FlashInfer
# BF16 (activation) x MXFP4 (weight) MoE backend.
"VLLM_USE_FLASHINFER_MOE_MXFP4_BF16": lambda: bool(
int(os.getenv("VLLM_USE_FLASHINFER_MOE_MXFP4_BF16", "0"))
# Deprecated: use --moe-backend to select a kernel explicitly.
"VLLM_USE_FLASHINFER_MOE_MXFP4_BF16": deprecated_env(
"VLLM_USE_FLASHINFER_MOE_MXFP4_BF16",
"v0.23",
"Use --moe-backend (e.g. flashinfer_trtllm, flashinfer_cutlass).",
lambda: bool(int(os.getenv("VLLM_USE_FLASHINFER_MOE_MXFP4_BF16", "0"))),
),
# Control the cache sized used by the xgrammar compiler. The default
# of 512 MB should be enough for roughly 1000 JSON schemas.
@@ -1480,10 +1536,17 @@ environment_variables: dict[str, Callable[[], Any]] = {
# Uses CUTLASS kernels optimized for high-throughput batch inference.
# - "latency":
# Uses TensorRT-LLM kernels optimized for low-latency inference.
"VLLM_FLASHINFER_MOE_BACKEND": env_with_choices(
# Deprecated: pass --moe-backend flashinfer_{trtllm,cutlass,cutedsl} directly.
"VLLM_FLASHINFER_MOE_BACKEND": deprecated_env(
"VLLM_FLASHINFER_MOE_BACKEND",
"latency",
["throughput", "latency", "masked_gemm"],
"v0.23",
"Use --moe-backend flashinfer_trtllm, flashinfer_cutlass, or "
"flashinfer_cutedsl.",
env_with_choices(
"VLLM_FLASHINFER_MOE_BACKEND",
"latency",
["throughput", "latency", "masked_gemm"],
),
),
# Override the directory for the FlashInfer autotune config cache.
"VLLM_FLASHINFER_AUTOTUNE_CACHE_DIR": lambda: os.getenv(
@@ -1565,8 +1628,12 @@ environment_variables: dict[str, Callable[[], Any]] = {
# Controls whether or not emulations are used for NVFP4
# generations on machines < 100 for compressed-tensors
# models
"VLLM_USE_NVFP4_CT_EMULATIONS": lambda: bool(
int(os.getenv("VLLM_USE_NVFP4_CT_EMULATIONS", "0"))
# Deprecated: use --linear-backend emulation instead.
"VLLM_USE_NVFP4_CT_EMULATIONS": deprecated_env(
"VLLM_USE_NVFP4_CT_EMULATIONS",
"v0.23",
"Use --linear-backend emulation.",
lambda: bool(int(os.getenv("VLLM_USE_NVFP4_CT_EMULATIONS", "0"))),
),
# Controls the read mode for the Mori-IO connector
"VLLM_MORIIO_CONNECTOR_READ_MODE": lambda: (
@@ -1601,18 +1668,24 @@ environment_variables: dict[str, Callable[[], Any]] = {
# This is only meant for research purposes to run on devices where NVFP4
# GEMM kernels are not available.
# - <none>: automatically pick an available backend
"VLLM_NVFP4_GEMM_BACKEND": env_with_choices(
# Deprecated: use --linear-backend instead.
"VLLM_NVFP4_GEMM_BACKEND": deprecated_env(
"VLLM_NVFP4_GEMM_BACKEND",
None,
[
"flashinfer-b12x",
"flashinfer-cudnn",
"flashinfer-trtllm",
"flashinfer-cutlass",
"cutlass",
"marlin",
"emulation",
],
"v0.23",
"Use --linear-backend.",
env_with_choices(
"VLLM_NVFP4_GEMM_BACKEND",
None,
[
"flashinfer-b12x",
"flashinfer-cudnn",
"flashinfer-trtllm",
"flashinfer-cutlass",
"cutlass",
"marlin",
"emulation",
],
),
),
# Controls garbage collection during CUDA graph capture.
# If set to 0 (default), enables GC freezing to speed up capture time.
@@ -1763,7 +1836,13 @@ environment_variables: dict[str, Callable[[], Any]] = {
# NCCL header path
"VLLM_NCCL_INCLUDE_PATH": lambda: os.environ.get("VLLM_NCCL_INCLUDE_PATH", None),
# Flag to enable FBGemm kernels on model execution
"VLLM_USE_FBGEMM": lambda: bool(int(os.getenv("VLLM_USE_FBGEMM", "0"))),
# Deprecated: use --linear-backend fbgemm instead.
"VLLM_USE_FBGEMM": deprecated_env(
"VLLM_USE_FBGEMM",
"v0.23",
"Use --linear-backend fbgemm.",
lambda: bool(int(os.getenv("VLLM_USE_FBGEMM", "0"))),
),
# GC debug config
# - VLLM_GC_DEBUG=0: disable GC debugger
# - VLLM_GC_DEBUG=1: enable GC debugger with gc.collect elpased times
vllm/model_executor/kernels/linear/__init__.py
+1 -21
View File
@@ -13,7 +13,6 @@ or kernel implementation, add it to this __init__.py to maintain
import stability.
"""
import warnings
from typing import TypeVar
import torch
@@ -851,31 +850,12 @@ def init_nvfp4_linear_kernel() -> NvFp4LinearKernel:
force_kernel = EmulationNvFp4LinearKernel
elif linear_backend == "auto":
# Deprecated env-var overrides — only honoured when --linear-backend
# is "auto". Will be removed in v0.21; users should migrate to
# --linear-backend.
# is "auto". Deprecation warnings are emitted from vllm/envs.py.
if envs.VLLM_USE_FBGEMM:
warnings.warn(
"VLLM_USE_FBGEMM is deprecated and will be removed in "
"v0.21. Use --linear-backend fbgemm instead.",
DeprecationWarning,
stacklevel=2,
)
force_kernel = FbgemmNvFp4LinearKernel
elif envs.VLLM_USE_NVFP4_CT_EMULATIONS:
warnings.warn(
"VLLM_USE_NVFP4_CT_EMULATIONS is deprecated and will be "
"removed in v0.21. Use --linear-backend emulation instead.",
DeprecationWarning,
stacklevel=2,
)
force_kernel = EmulationNvFp4LinearKernel
elif envs.VLLM_NVFP4_GEMM_BACKEND is not None:
warnings.warn(
"VLLM_NVFP4_GEMM_BACKEND is deprecated and will be "
"removed in v0.21. Use --linear-backend instead.",
DeprecationWarning,
stacklevel=2,
)
backend_name = envs.VLLM_NVFP4_GEMM_BACKEND
force_kernel = _NVFP4_BACKEND_TO_KERNEL.get(backend_name)
if force_kernel is None: