vladmandic / automatic

SD.Next: Advanced Implementation of Stable Diffusion and other Diffusion-based generative image models
https://github.com/vladmandic/automatic
GNU Affero General Public License v3.0
5.68k stars 422 forks source link

[Issue]: Can't set attribute 'dtype', when inpainting using SDXL #2452

Closed Silanda closed 12 months ago

Silanda commented 12 months ago

Issue Description

Since the latest update (as of 11th November), I'm getting this error when attempting to inpaint using an SDXL model.

Regular img2img works fine.

Version Platform Description

2023-11-07 13:17:02,769 | sd | INFO | launch | Starting SD.Next 2023-11-07 13:17:02,777 | sd | INFO | installer | Python 3.10.6 on Windows 2023-11-07 13:17:02,907 | sd | INFO | installer | Version: app=sd.next updated=2023-11-06 hash=a2b50b69 url=https://github.com/vladmandic/automatic.git/tree/master 2023-11-07 13:17:03,277 | sd | INFO | installer | Updating main repository 2023-11-07 13:17:05,132 | sd | INFO | installer | Upgraded to version: e53db4e2 Tue Nov 7 08:00:44 2023 -0500 2023-11-07 13:17:05,137 | sd | INFO | launch | Platform: arch=AMD64 cpu=AMD64 Family 23 Model 113 Stepping 0, AuthenticAMD system=Windows release=Windows-10-10.0.22621-SP0 python=3.10.6 2023-11-07 13:17:05,141 | sd | DEBUG | installer | Setting environment tuning 2023-11-07 13:17:05,141 | sd | DEBUG | installer | Torch overrides: cuda=False rocm=False ipex=False diml=False openvino=False 2023-11-07 13:17:05,141 | sd | DEBUG | installer | Torch allowed: cuda=True rocm=True ipex=True diml=True openvino=True 2023-11-07 13:17:05,147 | sd | INFO | installer | nVidia CUDA toolkit detected: nvidia-smi present 2023-11-07 13:17:05,227 | sd | INFO | launch | Startup: standard 2023-11-07 13:17:05,237 | sd | INFO | installer | Verifying requirements 2023-11-07 13:17:05,247 | sd | INFO | installer | Verifying packages 2023-11-07 13:17:05,252 | sd | INFO | installer | Verifying submodules 2023-11-07 13:17:09,237 | sd | DEBUG | installer | Submodule: extensions-builtin/sd-extension-chainner / main 2023-11-07 13:17:09,897 | sd | DEBUG | installer | Submodule: extensions-builtin/sd-extension-system-info / main 2023-11-07 13:17:10,457 | sd | DEBUG | installer | Submodule: extensions-builtin/sd-webui-agent-scheduler / main 2023-11-07 13:17:11,097 | sd | DEBUG | installer | Submodule: extensions-builtin/sd-webui-controlnet / main 2023-11-07 13:17:11,697 | sd | DEBUG | installer | Submodule: extensions-builtin/stable-diffusion-webui-images-browser / main 2023-11-07 13:17:12,297 | sd | DEBUG | installer | Submodule: extensions-builtin/stable-diffusion-webui-rembg / master 2023-11-07 13:17:12,959 | sd | DEBUG | installer | Submodule: modules/k-diffusion / master 2023-11-07 13:17:13,549 | sd | DEBUG | installer | Submodule: modules/lora / main 2023-11-07 13:17:14,127 | sd | DEBUG | installer | Submodule: wiki / master 2023-11-07 13:17:14,687 | sd | DEBUG | paths | Register paths 2023-11-07 13:17:14,927 | sd | DEBUG | installer | Installed packages: 216 2023-11-07 13:17:14,927 | sd | DEBUG | installer | Extensions all: ['clip-interrogator-ext', 'Lora', 'sd-extension-chainner', 'sd-extension-system-info', 'sd-webui-agent-scheduler', 'sd-webui-controlnet', 'stable-diffusion-webui-images-browser', 'stable-diffusion-webui-rembg'] 2023-11-07 13:17:14,982 | sd | DEBUG | installer | Submodule: extensions-builtin\clip-interrogator-ext / main 2023-11-07 13:17:15,528 | sd | DEBUG | installer | Running extension installer: I:\GitHub\automatic\extensions-builtin\clip-interrogator-ext\install.py 2023-11-07 13:17:24,187 | sd | DEBUG | installer | Submodule: extensions-builtin\sd-extension-chainner / main 2023-11-07 13:17:25,027 | sd | DEBUG | installer | Submodule: extensions-builtin\sd-extension-system-info / main 2023-11-07 13:17:25,527 | sd | DEBUG | installer | Running extension installer: I:\GitHub\automatic\extensions-builtin\sd-extension-system-info\install.py 2023-11-07 13:17:26,258 | sd | DEBUG | installer | Submodule: extensions-builtin\sd-webui-agent-scheduler / main 2023-11-07 13:17:26,843 | sd | DEBUG | installer | Running extension installer: I:\GitHub\automatic\extensions-builtin\sd-webui-agent-scheduler\install.py 2023-11-07 13:17:27,578 | sd | DEBUG | installer | Submodule: extensions-builtin\sd-webui-controlnet / main 2023-11-07 13:17:28,113 | sd | DEBUG | installer | Running extension installer: I:\GitHub\automatic\extensions-builtin\sd-webui-controlnet\install.py 2023-11-07 13:17:28,847 | sd | DEBUG | installer | Submodule: extensions-builtin\stable-diffusion-webui-images-browser / main 2023-11-07 13:17:29,387 | sd | DEBUG | installer | Running extension installer: I:\GitHub\automatic\extensions-builtin\stable-diffusion-webui-images-browser\install.py 2023-11-07 13:17:30,137 | sd | DEBUG | installer | Submodule: extensions-builtin\stable-diffusion-webui-rembg / master 2023-11-07 13:17:34,927 | sd | DEBUG | installer | Running extension installer: I:\GitHub\automatic\extensions-builtin\stable-diffusion-webui-rembg\install.py 2023-11-07 13:17:35,628 | sd | DEBUG | installer | Extensions all: ['canvas-zoom'] 2023-11-07 13:17:35,686 | sd | DEBUG | installer | Submodule: extensions\canvas-zoom / main 2023-11-07 13:17:36,212 | sd | DEBUG | installer | Running extension installer: I:\GitHub\automatic\extensions\canvas-zoom\install.py 2023-11-07 13:17:39,127 | sd | INFO | installer | Extensions enabled: ['clip-interrogator-ext', 'Lora', 'sd-extension-chainner', 'sd-extension-system-info', 'sd-webui-agent-scheduler', 'sd-webui-controlnet', 'stable-diffusion-webui-images-browser', 'stable-diffusion-webui-rembg', 'canvas-zoom'] 2023-11-07 13:17:39,127 | sd | INFO | installer | Verifying requirements 2023-11-07 13:17:39,148 | sd | INFO | installer | Updating Wiki 2023-11-07 13:17:39,198 | sd | DEBUG | installer | Submodule: i:\GitHub\automatic\wiki / master 2023-11-07 13:17:39,768 | sd | DEBUG | launch | Setup complete without errors: 1699363060 2023-11-07 13:17:39,768 | sd | INFO | installer | Extension preload: {'extensions-builtin': 0.0, 'extensions': 0.0} 2023-11-07 13:17:39,768 | sd | INFO | launch | Command line args: ['--upgrade'] upgrade=True 2023-11-07 13:17:47,498 | sd | INFO | loader | Load packages: torch=2.1.0+cu118 diffusers=0.22.0 gradio=3.43.2 2023-11-07 13:17:48,268 | sd | INFO | shared | Engine: backend=Backend.DIFFUSERS compute=cuda mode=no_grad device=cuda cross-optimization="Scaled-Dot-Product" 2023-11-07 13:17:48,318 | sd | INFO | shared | Device: device=NVIDIA GeForce RTX 4070 Ti n=1 arch=compute_37 cap=(8, 9) cuda=11.8 cudnn=8700 driver=546.01 2023-11-07 13:17:49,049 | sd | INFO | sd_vae | Available VAEs: path="models\VAE" items=2 2023-11-07 13:17:49,049 | sd | INFO | shared | Disabling uncompatible extensions: backend=Backend.DIFFUSERS ['sd-webui-controlnet', 'multidiffusion-upscaler-for-automatic1111', 'a1111-sd-webui-lycoris'] 2023-11-07 13:17:49,058 | sd | INFO | sd_models | Available models: path="models\Stable-diffusion" items=24 time=0.01 2023-11-07 13:17:51,591 | sd | INFO | script_loading | Extension: script='extensions-builtin\sd-webui-agent-scheduler\scripts\task_scheduler.py' Using sqlite file: extensions-builtin\sd-webui-agent-scheduler\task_scheduler.sqlite3 2023-11-07 13:17:52,648 | sd | INFO | webui | Extensions time: 3.33 { clip-interrogator-ext=1.52 Lora=0.20 sd-extension-chainner=0.08 sd-extension-system-info=0.05 sd-webui-agent-scheduler=0.39 stable-diffusion-webui-images-browser=0.12 stable-diffusion-webui-rembg=0.93 } 2023-11-07 13:17:52,928 | sd | INFO | theme | Load UI theme: name="black-orange" style=Auto base=style.css 2023-11-07 13:17:54,848 | sd | INFO | webui | Local URL: http://127.0.0.1:7860/ 2023-11-07 13:17:54,848 | sd | INFO | middleware | Initializing middleware 2023-11-07 13:17:55,063 | sd | INFO | task_runner | [AgentScheduler] Task queue is empty 2023-11-07 13:17:55,063 | sd | INFO | api | [AgentScheduler] Registering APIs 2023-11-07 13:17:55,168 | sd | INFO | sd_models | Select: model="pandorasBoxNSFW_v1PussBoots [a3cd261baa]" 2023-11-07 13:17:55,268 | sd | INFO | devices | Setting Torch parameters: device=cuda dtype=torch.float16 vae=torch.float16 unet=torch.float16 context=no_grad fp16=True bf16=False 2023-11-07 13:17:55,273 | sd | INFO | sd_vae | Loading VAE: model=models\VAE\sdxl_vae_fix source=settings 2023-11-07 13:17:55,751 | sd | INFO | sd_models | Autodetect: model="Stable Diffusion XL" class=StableDiffusionXLPipeline file="i:\GitHub\automatic\models\Stable-diffusion\pandorasBoxNSFW_v1PussBoots.safetensors" size=6617MB 2023-11-07 13:18:02,218 | sd | INFO | textual_inversion | Load embeddings: loaded=0 skipped=0 time=0.00 2023-11-07 13:18:02,455 | sd | INFO | sd_models | Load model: time=7.05 { load=7.05 } native=512 {'ram': {'used': 8.18, 'total': 31.93}, 'gpu': {'used': 8.01, 'total': 11.99}, 'retries': 0, 'oom': 0} 2023-11-07 13:18:02,458 | sd | INFO | webui | Startup time: 22.68 { torch=6.92 gradio=0.74 diffusers=0.06 libraries=1.54 extensions=3.33 face-restore=0.25 upscalers=0.14 extra-networks=0.06 ui-extra-networks=0.21 ui-img2img=0.06 ui-settings=0.14 ui-extensions=1.04 ui-defaults=0.26 launch=0.22 api=0.08 app-started=0.23 checkpoint=7.29 }

Relevant log output

2023-11-07 13:29:14,112 | sd | ERROR | call_queue | Exception: can't set attribute 'dtype'
2023-11-07 13:29:14,113 | sd | ERROR | call_queue | Arguments: args=('task(2n91zm1pjrssjkp)', 2, 'cyberpunk painting, painterly, extremely detailed, a beautiful nude robotic young woman with beautiful eyes and small breasts in bizarre biomechanical setting, arms crossed, metal skin, geometric tattoos, beautiful nipples, long hair, weird alien buildings in background, flying craft in sky', 'asymmetric eyes, large forehead', [], None, None, {'image': <PIL.Image.Image image mode=RGBA size=1344x768 at 0x1B77FC8D600>, 'mask': <PIL.Image.Image image mode=RGB size=1344x768 at 0x1B77FC8FFD0>}, None, None, None, None, 10, 1, 13, 32, 0, 1, True, False, False, 1, 1, 8, 1.5, 0.7, 10, 0, 1, 0.2, -1.0, -1.0, 0, 0, 0, 0, 1280, 1280, 1, 0, 1, 32, 0, None, '', '', '', [], 0, '* `CFG Scale` should be 2 or lower.', True, True, '', '', True, 50, True, 1, 0, False, 4, 0.5, 'Linear', 'None', '<p style="margin-bottom:0.75em">Recommended settings: Sampling Steps: 80-100, Sampler: Euler a, Denoising strength: 0.8</p>', 128, 8, ['left', 'right', 'up', 'down'], 1, 0.05, 128, 4, 0, ['left', 'right', 'up', 'down'], False, False, 'positive', 'comma', 0, False, False, '', '<p style="margin-bottom:0.75em">Will upscale the image by the selected scale factor; use width and height sliders to set tile size</p>', 256, 30, 3, 0, '', [], 0, '', [], 0, '', [], False, True, False, False, False, False, 0) kwargs={}
2023-11-07 13:29:14,118 | sd | ERROR | errors | gradio call: AttributeError

Backend

Diffusers

Model

SD-XL

Acknowledgements

vladmandic commented 12 months ago

fixed.

Silanda commented 12 months ago

Thanks!