vladmandic
commented
2 hours ago never had anything like that with other SD software. Even enabling Nvidia's system fallback doesn't help, this is ridiculous. How do I configure it so that it doesn't perform worse than others?
i can understand the frustration, but i still do not like the entitled tone. above anything, this is free and open source and relies on its community more than anything.
now, if you actually want to change the tone and proceed with troubleshooting, run with --debug flag and upload a full log here as i cannot see params used when oom happened.
Issue Description
I installed it via Pinokio, and running anything above 512x512 batch 1 does OOM at my 8Gb 2070S, I never had anything like that with other SD software. Even enabling Nvidia's system fallback doesn't help, this is ridiculous. How do I configure it so that it doesn't perform worse than others? I went for SD Next hoping I'd be able to run modern models that Forge can't and here I'm left unable to use even the oldest one. There are way too many settings to brute-force through their combinations on my own.
Version Platform Description
Win10 x64, 20GB RAM. 20:58:26-298795 INFO Starting SD.Next 20:58:26-308950 INFO Logger: file="f:\pinokio\api\SD-Next.git\app\sdnext.log" level=INFO size=97458 mode=append 20:58:26-308950 INFO Python: version=3.10.15 platform=Windows bin="f:\pinokio\api\SD-Next.git\app\venv\Scripts\python.exe" venv="f:\pinokio\api\SD-Next.git\app\venv" 20:58:26-488706 INFO Version: app=sd.next updated=2024-11-02 hash=65ddc611 branch=master url=https://github.com/vladmandic/automatic/tree/master ui=main 20:58:27-278944 INFO Platform: arch=AMD64 cpu=Intel64 Family 6 Model 42 Stepping 7, GenuineIntel system=Windows release=Windows-10-10.0.19045-SP0 python=3.10.15 20:58:27-288602 INFO Args: ['--use-cuda'] 20:58:27-333397 INFO CUDA: nVidia toolkit detected 20:58:27-598652 INFO Verifying requirements 20:58:27-608594 INFO Verifying packages 20:58:27-668815 INFO Extensions: disabled=[] 20:58:27-668815 INFO Extensions: enabled=['Lora', 'sd-extension-chainner', 'sd-extension-system-info', 'sd-webui-agent-scheduler', 'sdnext-modernui', 'stable-diffusion-webui-rembg'] extensions-builtin 20:58:27-678921 INFO Extensions: enabled=[] extensions 20:58:27-678921 INFO Startup: quick launch 20:58:27-678921 INFO Extensions: disabled=[] 20:58:27-678921 INFO Extensions: enabled=['Lora', 'sd-extension-chainner', 'sd-extension-system-info', 'sd-webui-agent-scheduler', 'sdnext-modernui', 'stable-diffusion-webui-rembg'] extensions-builtin 20:58:27-688612 INFO Extensions: enabled=[] extensions 20:58:27-698470 INFO Command line args: ['--use-cuda'] use_cuda=True 20:58:38-309021 INFO System packages: {'torch': '2.5.1+cu124', 'diffusers': '0.32.0.dev0', 'gradio': '3.43.2', 'transformers': '4.46.1', 'accelerate': '1.0.1'} 20:58:39-498829 INFO Device detect: memory=8.0 optimization=medvram 20:58:39-509009 INFO Engine: backend=Backend.DIFFUSERS compute=cuda device=cuda attention="Scaled-Dot-Product" mode=no_grad 20:58:39-569152 INFO Torch parameters: backend=cuda device=cuda config=Auto dtype=torch.bfloat16 vae=torch.bfloat16 unet=torch.bfloat16 context=no_grad nohalf=False nohalfvae=False upscast=False deterministic=False test-fp16=True test-bf16=True optimization="Scaled-Dot-Product" 20:58:40-428618 INFO Device: device=NVIDIA GeForce RTX 2070 SUPER n=1 arch=sm_90 capability=(7, 5) cuda=12.4 cudnn=90100 driver=560.81 20:58:40-678641 INFO Available VAEs: path="models\VAE" items=0 20:58:40-678641 INFO Available UNets: path="models\UNET" items=0 20:58:40-678641 INFO Available TEs: path="models\Text-encoder" items=0 20:58:40-678641 INFO Disabled extensions: ['sdnext-modernui'] 20:58:40-708635 INFO Available Models: path="models\Stable-diffusion" items=30 time=0.02 20:58:40-828603 INFO Available Yolo: path="models\yolo" items=6 downloaded=0 20:58:40-968287 INFO Available LoRAs: path="models\Lora" items=0 folders=2 time=0.00 20:58:42-178467 INFO Extension: script='extensions-builtin\sd-webui-agent-scheduler\scripts\task_scheduler.py' Using sqlite file: extensions-builtin\sd-webui-agent-scheduler\task_scheduler.sqlite3 20:58:42-218510 INFO Available Upscalers: items=53 downloaded=0 user=0 time=0.04 types=['None', 'Lanczos', 'Nearest', 'ChaiNNer', 'AuraSR', 'ESRGAN', 'LDSR', 'RealESRGAN', 'SCUNet', 'SD', 'SwinIR'] 20:58:42-248951 INFO Available Styles: folder="models\styles" items=288 time=0.03 20:58:42-258470 INFO UI theme: type=Standard name="black-teal" 20:58:44-779036 INFO Extension list is empty: refresh required 20:58:45-958965 INFO Local URL: http://127.0.0.1:7860/ 20:58:46-409037 INFO [AgentScheduler] Task queue is empty 20:58:46-409037 INFO [AgentScheduler] Registering APIs 20:58:46-648893 INFO Load model: select="stable-diffusion!sd-1-4 [764ebd128c]" 20:58:46-658906 INFO Autodetect model: detect="Stable Diffusion" class=StableDiffusionPipeline file="f:\pinokio\api\SD-Next.git\app\models\Stable-diffusion\stable-diffusion!sd-1-4.safetensors" size=2034MB Diffusers 3.19it/s ████ 50% 3/6 00:00 00:00 Loading pipeline components... Diffusers 1.30s/it ████████ 100% 6/6 00:07 00:00 Loading pipeline components... 20:58:54-790835 INFO Load network: type=embeddings loaded=0 skipped=0 time=0.00 20:58:55-128450 INFO Load model: time=8.19 load=8.13 native=512 memory={'ram': {'used': 2.98, 'total': 19.97}, 'gpu': {'used': 1.06, 'total': 8.0}, 'retries': 0, 'oom': 0} 20:58:55-128450 INFO Startup time: 27.43 torch=8.02 onnx=0.05 gradio=2.05 diffusers=0.23 libraries=2.55 samplers=0.05 extensions=1.34 detailer=0.13 ui-networks=0.36 ui-txt2img=0.36 ui-img2img=0.20 ui-control=0.38 ui-extras=0.10 ui-models=0.12
ui-gallery=0.08 ui-settings=0.88 ui-extensions=0.58 ui-defaults=0.38 launch=0.28 api=0.27 app-started=0.41 checkpoint=8.48 20:59:39-234929 INFO MOTD: N/A 20:59:45-526194 INFO Browser session: user=None client=127.0.0.1 agent=Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Pinokio/2.15.1 Chrome/110.0.5481.208 Electron/23.3.13 Safari/537.36
Relevant log output
Backend
Diffusers
UI
Standard
Branch
Master
Model
StableDiffusion 1.5
Acknowledgements