[X] I have searched the existing issues and checked the recent builds/commits
What happened?
I got RuntimeWarning: invalid value encountered in cast x_sample = x_sample.astype(np.uint8) trying to use a lora trained by https://github.com/bmaltais/kohya_ss which is also the newest version v21.8.7 with these settings https://www.mediafire.com/file/p7hc4q8n7wuu31z/CharacterLoraSettings.json/file. The output in the web ui is also just black too but the samples from kohya look fine. Any other lora works just fine and dose not have that error or issue, just the ones I train.
Steps to reproduce the problem
Go to ....
Press ....
...
load my lora
tell it to produce an image
What should have happened?
It should have not given me that error or just a black output.
Version or Commit where the problem happens
1.5.1
What Python version are you running on ?
Python 3.10.x
What platforms do you use to access the UI ?
Windows
What device are you running WebUI on?
Nvidia GPUs (RTX 20 above)
Cross attention optimization
xformers
What browsers do you use to access the UI ?
Google Chrome
Command Line Arguments
No
List of extensions
Yes
sd_dreambooth_extension
Console logs
Using VENV: D:\stable-diffusion2\automatic\venv
08:44:39-663592 INFO Starting SD.Next
08:44:39-683542 INFO Python 3.10.11 on Windows
08:44:39-775292 INFO Version: 1666ee68 Sat Aug 19 19:44:58 2023 +0300
08:44:41-207729 INFO Latest published version: 70395fb63a76f99579f040e67f3c55ad17f6035b 2023-08-20T12:16:06Z
08:44:41-212715 INFO nVidia CUDA toolkit detected
08:44:41-609698 INFO Disabled extensions: []
08:44:41-611645 INFO Enabled extensions-builtin: ['a1111-sd-webui-lycoris', 'clip-interrogator-ext', 'LDSR',
'Lora', 'multidiffusion-upscaler-for-automatic1111', 'ScuNET', 'sd-dynamic-thresholding',
'sd-extension-system-info', 'sd-webui-agent-scheduler', 'sd-webui-controlnet',
'stable-diffusion-webui-images-browser', 'stable-diffusion-webui-rembg', 'SwinIR']
08:44:41-621622 INFO Enabled extensions: ['sd_dreambooth_extension']
08:44:41-624616 INFO No changes detected: Quick launch active
08:44:41-625608 INFO Verifying requirements
08:44:41-652537 INFO Verifying packages
08:44:41-656530 INFO Disabled extensions: []
08:44:41-658520 INFO Enabled extensions-builtin: ['a1111-sd-webui-lycoris', 'clip-interrogator-ext', 'LDSR',
'Lora', 'multidiffusion-upscaler-for-automatic1111', 'ScuNET', 'sd-dynamic-thresholding',
'sd-extension-system-info', 'sd-webui-agent-scheduler', 'sd-webui-controlnet',
'stable-diffusion-webui-images-browser', 'stable-diffusion-webui-rembg', 'SwinIR']
08:44:41-667500 INFO Enabled extensions: ['sd_dreambooth_extension']
08:44:41-751276 INFO Extension preload: 0.0s D:\stable-diffusion2\automatic\extensions-builtin
08:44:41-772219 INFO Extension preload: 0.0s D:\stable-diffusion2\automatic\extensions
08:44:41-847021 INFO Server arguments: []
08:45:57-633734 INFO Pipeline: Backend.ORIGINAL
08:46:03-507649 INFO Libraries loaded
08:46:03-510639 INFO Using data path: D:\stable-diffusion2\automatic
08:46:03-514634 INFO Available VAEs: D:\stable-diffusion2\automatic\models\VAE 0
08:46:03-554522 INFO Available models: D:\stable-diffusion2\automatic\models\Stable-diffusion 3
08:46:10-457457 INFO ControlNet v1.1.234
ControlNet v1.1.234
ControlNet preprocessor location: D:\stable-diffusion2\automatic\extensions-builtin\sd-webui-controlnet\annotator\downloads
08:46:10-875785 INFO ControlNet v1.1.234
ControlNet v1.1.234
08:46:15-107524 WARNING Selected checkpoint not found: CharacterLora-000014.safetensors [ac70463d5a]
Selected checkpoint not found: CharacterLora-000014.safetensors [ac70463d5a]
Loading weights: D:\stable-diffusion2\automatic\models\Stable-diffusion\CharacterLora.safetensors ------ 0.0/1… -:--:--
MB
08:46:15-488502 INFO Torch override dtype: no-half set
Torch override dtype: no-half set
08:46:15-489503 INFO Torch override VAE dtype: no-half set
Torch override VAE dtype: no-half set
08:46:15-490496 INFO Setting Torch parameters: dtype=torch.float32 vae=torch.float32 unet=torch.float32
Setting Torch parameters: dtype=torch.float32 vae=torch.float32 unet=torch.float32
LatentDiffusion: Running in eps-prediction mode
DiffusionWrapper has 859.52 M params.
08:46:28-955953 INFO Applying xformers cross attention optimization
Applying xformers cross attention optimization
08:46:29-005852 INFO Embeddings: loaded=0 skipped=0
Embeddings: loaded=0 skipped=0
08:46:29-013802 INFO Model loaded in 13.7s (load=0.2s create=11.6s move=1.9s)
Model loaded in 13.7s (load=0.2s create=11.6s move=1.9s)
08:46:29-221325 INFO Model load finished: {'ram': {'used': 1.02, 'total': 15.77}, 'gpu': {'used': 5.11, 'total':
12.0}, 'retries': 0, 'oom': 0} cached=0
Model load finished: {'ram': {'used': 1.02, 'total': 15.77}, 'gpu': {'used': 5.11, 'total': 12.0}, 'retries': 0, 'oom': 0} cached=0
08:46:30-368442 INFO Loading UI theme: name=black-orange style=Auto
08:46:31-091116 INFO Available models: D:\stable-diffusion2\automatic\models\Stable-diffusion 3
Running on local URL: http://127.0.0.1:7860
08:46:32-577922 INFO Local URL: http://127.0.0.1:7860/
08:46:32-580881 INFO Initializing middleware
08:46:32-747437 INFO [AgentScheduler] Task queue is empty
08:46:32-750428 INFO [AgentScheduler] Registering APIs
08:46:32-859825 INFO Startup time: 111.0s (torch=55.8s gradio=7.0s libraries=18.8s codeformer=0.9s gfpgan=0.4s
scripts=24.9s onchange=0.6s ui-txt2img=0.1s ui-img2img=0.1s ui-settings=0.1s
ui-extensions=1.5s ui-defaults=0.1s launch=0.4s app-started=0.3s)
D:\stable-diffusion2\automatic\modules\sd_samplers_common.py:42: RuntimeWarning: invalid value encountered in cast
x_sample = np.moveaxis(x_sample.numpy(), 0, 2).astype(np.uint8)
Progress 2.28it/s ---------------------------------------- 100% 0:00:00 0:00:08
D:\stable-diffusion2\automatic\modules\processing.py:733: RuntimeWarning: invalid value encountered in cast
x_sample = x_sample.astype(np.uint8)
Is there an existing issue for this?
What happened?
I got RuntimeWarning: invalid value encountered in cast x_sample = x_sample.astype(np.uint8) trying to use a lora trained by https://github.com/bmaltais/kohya_ss which is also the newest version v21.8.7 with these settings https://www.mediafire.com/file/p7hc4q8n7wuu31z/CharacterLoraSettings.json/file. The output in the web ui is also just black too but the samples from kohya look fine. Any other lora works just fine and dose not have that error or issue, just the ones I train.
Steps to reproduce the problem
What should have happened?
It should have not given me that error or just a black output.
Version or Commit where the problem happens
1.5.1
What Python version are you running on ?
Python 3.10.x
What platforms do you use to access the UI ?
Windows
What device are you running WebUI on?
Nvidia GPUs (RTX 20 above)
Cross attention optimization
xformers
What browsers do you use to access the UI ?
Google Chrome
Command Line Arguments
List of extensions
Yes
sd_dreambooth_extension
Console logs
Additional information
No response