Skip to content

[BugFix] Fix noisy output without setting a seed in Qwen Image #1043

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
hsliuustc0106 merged 1 commit into from
Jan 29, 2026
Merged

[BugFix] Fix noisy output without setting a seed in Qwen Image #1043

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 8 additions & 0 deletions vllm_omni/entrypoints/openai/api_server.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,7 @@
import os

# Image generation API imports
import random
import time
from argparse import Namespace
from collections.abc import AsyncIterator
Expand Down Expand Up @@ -984,6 +985,13 @@ async def generate_images(request: ImageGenerationRequest, raw_request: Request)
gen_params.true_cfg_scale = request.true_cfg_scale
if request.seed is not None:
gen_params.seed = request.seed
else:
# If seed is not provided, generate a random one to ensure
# a proper generator is initialized in the backend.
# This fixes issues where using the default global generator
# might produce blurry images in some environments.
gen_params.seed = random.randint(0, 2**32 - 1)

request_id = f"img_gen_{int(time.time())}"

logger.info(f"Generating {request.n} image(s) {size_str}")
Expand Down