This video looks at how to run Stable Diffusion on Run Pod. Connect with me on Instagram / alphonso_peluso and TikTok / alphonso_peluso #stablediffusion #controlnet #runpod
thank you sir. In my case, I've chosen RunPod Fast SD because it has comfyUI as well but when I terminate and redo the process I do not have the sd folder thus need to 'run all cells' to see the relevant folders again in Jupyter. What do I need to do?
@@AlphonsoPeluso Thank you. I didnt use the community one. I think maybe I need to save teh workbook or notebook? The interface is now slightly different from your video.
Brilliant tutorial! Thank you so much. When I try and load the stabel diffusion img2vid model after connecting through the HTTP port, the model doesn't change after I select it from the drop down menu. Any idea why this might be the case? Does this RunPod configuration not support image to video yet?
Thanks a lot for the intutive tutorial. I'd like to check if you want to start with the same stable diffusion again on the same server, how should we do that? Should we wait for the same period of time (more than 10 mins according to your video) for the server to deply everything again every time when we want to use it? Thanks ;)
You’re welcome. If you use a Secure server with a Network Volume drive the pod will start up much faster and all your files will be on the network volume drive.
@@AlphonsoPeluso it's working now, it said "HTTP request sent 401 unauthorized authentification failed" don't know what caused that but 20 min later it's all good now, only problem is the Lora isn't showing up in the program even though it's in the folder, checkpoint shows up no problem, getting ready to give up.
when i try to copy and paste my checkpoint i get this msg how to fix it ? File Load Error for Ohara.safetensors /workspace/stable-diffusion-webui/models/Stable-diffusion/Ohara.safetensors is not UTF-8 encoded
Thanks! It was really helpful because the difference between a community/secure server was unclear in another video, and I was looking for information on how to save my models and Loras after terminating the Pod. Runpod, in my opinion, is more convenient than Google Colab, but sometimes the waiting time to set up the server is too long.
alphonso with the architectural model it said unauthorized 401.
whats that? Thank you. 😊
@@content1 Try the older version. The new one doesn’t work
thank you sir. In my case, I've chosen RunPod Fast SD because it has comfyUI as well but when I terminate and redo the process I do not have the sd folder thus need to 'run all cells' to see the relevant folders again in Jupyter. What do I need to do?
You need to setup a network storage volume drive to store your files so you can re access them.
Alphonso this was a great tutorial. You just saved me 4000 dollars in a computer. I will use this system. 🎉🎉🎉 THANK YOU!
@@content1 Lol. Nice to hear it! Thanks for the feedback.
Alphonso what I am doing wrong as it didnt save the models neither the pod for next day?
@@content1 You most likely used a community cloud server. You need to use a secure cloud server with a network volume drive
@@AlphonsoPeluso Thank you. I didnt use the community one. I think maybe I need to save teh workbook or notebook? The interface is now slightly different from your video.
very helpful thank you!
@@des_bloom You’re welcome!
Thank you. I am not tech saavy and have been suffering with this a couple days. Got it now. Thanks again
😀 I’m glad you got it. Have fun!
Perfect! Finally someone who speaks English of the commons!
lol. 😂 I’m glad you liked it!
exactly!
@@redampsmining lol 😂
Brilliant tutorial! Thank you so much. When I try and load the stabel diffusion img2vid model after connecting through the HTTP port, the model doesn't change after I select it from the drop down menu. Any idea why this might be the case? Does this RunPod configuration not support image to video yet?
Hello. Are you sure that you have the model in the correct folder? Make sure that your under workspace.
thx for a great video!
@@kasperrzuelow You’re welcome. Thanks for watching 🙌🔥🙌
Thanks a lot for the intutive tutorial. I'd like to check if you want to start with the same stable diffusion again on the same server, how should we do that? Should we wait for the same period of time (more than 10 mins according to your video) for the server to deply everything again every time when we want to use it? Thanks ;)
You’re welcome. If you use a Secure server with a Network Volume drive the pod will start up much faster and all your files will be on the network volume drive.
Love Please carry on with further vedios
@@arumairdesigns682 Thank you 🙏
What about creating an API endpoint and running it on runpod
🤷♂️
Can you do a video on how to save the extensions, when I delete the pod, I have to start for scratch again
You can’t use a community cloud server. You need a secure cloud server with a volume network drive to store your files
@@AlphonsoPeluso Yeah I tried that Ill try again - Its the saving the files part I'm having issues with
good video. When I tried to upload the Lora with wget it didn't work.
You need to make sure the web address doesn’t have any extra text.
@@AlphonsoPeluso it's working now, it said "HTTP request sent 401 unauthorized authentification failed" don't know what caused that but 20 min later it's all good now, only problem is the Lora isn't showing up in the program even though it's in the folder, checkpoint shows up no problem, getting ready to give up.
@@starg47 Did you make sure to save it with a .safetensors extension?
@@AlphonsoPeluso yeah, it's working now, some loras don't work, just tried a different one.
when i try to copy and paste my checkpoint i get this msg how to fix it ?
File Load Error for Ohara.safetensors
/workspace/stable-diffusion-webui/models/Stable-diffusion/Ohara.safetensors is not UTF-8 encoded
Thanks! It was really helpful because the difference between a community/secure server was unclear in another video, and I was looking for information on how to save my models and Loras after terminating the Pod. Runpod, in my opinion, is more convenient than Google Colab, but sometimes the waiting time to set up the server is too long.
Yes! I agree with you 💯Thanks for the comment and feedback. I wish you all the best.