From 46fae69cf04db85a3b187a5fdc061b21e24b9571 Mon Sep 17 00:00:00 2001 From: Cyrus Leung Date: Sun, 4 May 2025 06:59:24 +0800 Subject: [PATCH] [Misc] V0 fallback for `--enable-prompt-embeds` (#17615) Signed-off-by: DarkLight1337 --- vllm/engine/arg_utils.py | 6 ++++++ vllm/inputs/preprocess.py | 3 --- 2 files changed, 6 insertions(+), 3 deletions(-) diff --git a/vllm/engine/arg_utils.py b/vllm/engine/arg_utils.py index f6f8fb69fb706..08dbb4c450393 100644 --- a/vllm/engine/arg_utils.py +++ b/vllm/engine/arg_utils.py @@ -1237,6 +1237,12 @@ class EngineArgs: recommend_to_remove=False) return False + # No text embedding inputs so far. + if self.enable_prompt_embeds: + _raise_or_fallback(feature_name="--enable-prompt-embeds", + recommend_to_remove=False) + return False + # Only Fp16 and Bf16 dtypes since we only support FA. V1_SUPPORTED_DTYPES = [torch.bfloat16, torch.float16] if model_config.dtype not in V1_SUPPORTED_DTYPES: diff --git a/vllm/inputs/preprocess.py b/vllm/inputs/preprocess.py index 53e0a477a12df..fe4775b210a87 100644 --- a/vllm/inputs/preprocess.py +++ b/vllm/inputs/preprocess.py @@ -6,7 +6,6 @@ from typing import Any, Optional, Union, cast from typing_extensions import assert_never -from vllm import envs from vllm.config import ModelConfig from vllm.logger import init_logger from vllm.lora.request import LoRARequest @@ -306,8 +305,6 @@ class InputPreprocessor: if not self.model_config.enable_prompt_embeds: raise ValueError("You must set `--enable-prompt-embeds` to input " "`prompt_embeds`.") - if envs.VLLM_USE_V1: - raise ValueError("`prompt_embeds` is only available in V0.") prompt_embeds = parsed_content["prompt_embeds"]