Today we dive into an Easy guide for installing ENTIRELY Local AI Video Generation for FREE on your machine at home! HUGE Thanks to Gigabyte @AORUS for sponsoring today's video! *GIGABYTE LAPTOP GIVEAWAY:* www.gigabyte.com/us/unleash-ai-creativity/ *Shop AORUS laptops:* aorus.shop/MattVidPro-AI ▼ Link(s) From Today’s Video: Pinokio: pinokio.computer/
Matt, your channel is my favorite, and ever since I joined the discord ive had endless fun and ideas. You have such a great community man. I'd love to see you interact with the discord more, especially in the ai creation sharing and dalle and bing creator. Have a great day!
@@MattVidPro Probably already fixed the issue by now, but in case you haven't. Most microphones only record in mono and OBS's default is stereo audio capture. Just go to the advanced audio properties in OBS audio mixer and set your microphone input to mono. You can also fix it in post if your mic track is recorded separately by setting that audio track to mono!
Thanks for the cool video, it was a lot easier to set up than I had expected based on what I read a year or so ago, so I'll definitely be trying it out and seeing if my old gpu can handle it ok. (: It's really uncanny how much AI's been improving even just in the past year and how much has been available to the public and not monopolized by big corps. I feel that the coming year or 2 is going to be extremely interesting with all the advancements that will be made and hopefully it will also become available to the public and be viable on consumer hardware. And thanks to both the channel and gigabyte for the giveaway for the AORUS 17X! I hope whoever wins it has an excellent time with it in games and for AI stuff!
I like the positive and uplifting vibes i get from ya and yes from a single comment lol but this is beyond facts and it wasn’t just AI Software lmaoo i remember my 12 year old self having false hope and thinking ANY gpu makes any game work so the day i was able to get some cheese gathered up and copped a cheap low profile geforce 210? Don’t remember specifically but only remember it being cheap and horrendous but i swear every game i would “try” and just get passed the menu i took it as a “ im on pc master race status with my 15-20 fps in game” and even after going below any minimum video settings if it was possible💀💀
Pretty cool that you can generate clips like that on your local system now, though the technology clearly has plenty of limitations and there's lots of room for future improvement.
I attempted to trim this video to see if UA-cam's video editor rerenders it in mono. it seems that there was an issue with the mono audio recording being converted to stereo on accident, my apologies.
@@MattVidPro I'm praying for a K.I. tool to come out soon that can perfect audio stereo distribution - Little joke, you're the best :) Wink 😉, this comment was crafted with a little help from ChatGPT
I've been using open-source AI stuff, like XTTS, Fooocus, and this one. Yes, it can be a hassle installing open-source things and waiting for it to be installed. But if you have a good PC you can do many things that other are paying for, and that is a win in my book. Plant the seeds of struggle, reap the harvest of ease.
Having AI being this freely accessible will be both dangerous and positive. It lowers the barrier of needing expertise but with that comes both good people being able to make good things and bad people making bad things.
Big problem with this is it doesn't tell you the size of the installs. Not good if you're trying to manage space, some of the installs are quite large and you have no idea how much of a dent they'll make until you've installed them.
I hope to try this technology, I have only used Windows 7 and Chromebook computers, I have never even seen the later and evolved versions of computers and PCs.
Nice and very easy to follow even for someone who has nearly no experience with computers so far. I would have preferred to see directly what the hardware requirements are, not just looking at task manager and hearing chances are pretty good if... Keep up the good work!
Hey guys! If you get the error “streamlit’ is not recognized as an internal or external command, operable program or batch file” perhaps the streamlit module is not integrated into python, pip install streamlit.
Awesome video, very informative thank you for your wisdom and teaching! AI is blowing up! It's truly amazing what AI can do and what you can do with AI and this is only the beginning!
It's kind of disappointing that it has such a limited time generation. The Forge version of SVD allows you to adjust the clip length. Maybe I'm missing something but I couldn't get anything better than 1 second.
Installation of ffmpeg failed at first, I had to update my conda install, after that it worked. I already had an old version of Pinokio, that yielded so many errors, that I gave up back then. Maybe it was the old conda version all along 🤔 Thanks Matt 👍
I'm intrigued by its potential for AI video generation. The powerful specs, including the GeForce RTX 4090 GPU and 13th Gen Intel i9 processor, sound amazing.
Nice simple tutorial thanks for the video. AI is going to have a greater impact on all aspects of our lives as it improves and we find new ways to utilize it.
Has anyone received this error? "LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on." I can see that it exists locally. But it isn't able to access it.
@Gafferman it probably puts some meta data in but you could remove it. But if you really making content you'd probably just use a paid service eventually.
i got virus alerts from some of the pinokio installs. Any chance to run this on a VM ? Another issue is that it installs versions of programs that can create issues with other AI tools.
I had a play on this for a few hours, was getting poor results. I found the clips too short and if u extended them the image would fall apart. Maybe some1 out there does a deep dive into getting great results.
to be fair, Stable diffusion is known for crap base models. The only reason it's so popular and widely known is the community that improves and build upon it. It wouldn't be insane to see in 1 year, 'models' for specific use cases so that way depending on what you want to do, you'll get better results based on that, (anime,people,animals,objects,backgrounds,etc) then you combine all models into one similar to SD image models. And probably the end result would be even better stuff than what the paid services would offer.
@@harnageaa seems like a hella lot of sliders and parameters. Does my head in all that, you need effective simple solutions. If your trying to make money you cant keep hoping and praying for tech and then theres learning and finding great reusable settings. As it stands comfy ui & sd will never take off because no one wants to deal with all the $hit that comes along with it. Eg : imagine how much wiring a record studio has. All the mics for drums, guitar, bass, keyboards, vox. They then go to amps, compressors, fx equipment... then into a 32 channel mixing desk ... I own a DAW (digital audio workstation). I simply record my audio & drop in fx. These DAWs sell in their millions. Same with photoshop, blender etc. The premium companies like midjourney will always be at forefront because they look to improve and simplify to make customers happy.
The only reason I am praising "open source models", is not because the models and the company offers an easy solution. I will give a direct example with image generation. After SD came out, there are now a TON of websites which took those "SD models" made themselves custom, and created AI PRODUCTS. Same for LLMs, they took the "free open source" models, tweaked, and made PRODUCTS. What I care is not the RAW open source interface, but community going in, creating stuff, then others creating products with the results of the community. Comfy ui is worst ui , in my opinion. It might have appeal for others but not for me. I'm more of a rectangle and sliders type of guy with tabs in it. @@armondtanz
Are you sure the "T" field in the left-hand sidebar doesn't stand for "Time"? It makes sense to me that if you set T=14 and FPS=14, you get 1 second of output (14 frames at 14fps).
Honestly I'm a bit worried at how well some AI generated videos look already... It probably won't take long before they'll become indistinguishable from real videos 😨
Today we dive into an Easy guide for installing ENTIRELY Local AI Video Generation for FREE on your machine at home! HUGE Thanks to Gigabyte @AORUS for sponsoring today's video!
*GIGABYTE LAPTOP GIVEAWAY:* www.gigabyte.com/us/unleash-ai-creativity/
*Shop AORUS laptops:* aorus.shop/MattVidPro-AI
▼ Link(s) From Today’s Video:
Pinokio: pinokio.computer/
👋
@@LouisGedowhats uppp
@@MattVidPro 👍
So which ai should I use for an amd gpu? Thanks!
My left ear enjoyed this video lmao, love you though Matt, I know you put so much work in every video
Many apologies and thank you for the kind words, I am attempting to fix it inside UA-cam!!
@@MattVidPro No problem
Matt, your channel is my favorite, and ever since I joined the discord ive had endless fun and ideas. You have such a great community man. I'd love to see you interact with the discord more, especially in the ai creation sharing and dalle and bing creator. Have a great day!
Thank you so much for the kind words!!! I also love the discord community! I wish I could spend more time there! I try and check in everyday!
my left ear really enjoyed this video, cheers matt!
I am so sorry about that!!! Working to try and fix it rn
@MattVidPro No worries man, thanks for the quality content as always!
I can tell it was edited on a Mac. XD
@@MattVidPro Probably already fixed the issue by now, but in case you haven't. Most microphones only record in mono and OBS's default is stereo audio capture. Just go to the advanced audio properties in OBS audio mixer and set your microphone input to mono. You can also fix it in post if your mic track is recorded separately by setting that audio track to mono!
Thanks for the cool video, it was a lot easier to set up than I had expected based on what I read a year or so ago, so I'll definitely be trying it out and seeing if my old gpu can handle it ok. (: It's really uncanny how much AI's been improving even just in the past year and how much has been available to the public and not monopolized by big corps. I feel that the coming year or 2 is going to be extremely interesting with all the advancements that will be made and hopefully it will also become available to the public and be viable on consumer hardware. And thanks to both the channel and gigabyte for the giveaway for the AORUS 17X! I hope whoever wins it has an excellent time with it in games and for AI stuff!
I like the positive and uplifting vibes i get from ya and yes from a single comment lol but this is beyond facts and it wasn’t just AI Software lmaoo i remember my 12 year old self having false hope and thinking ANY gpu makes any game work so the day i was able to get some cheese gathered up and copped a cheap low profile geforce 210? Don’t remember specifically but only remember it being cheap and horrendous but i swear every game i would “try” and just get passed the menu i took it as a “ im on pc master race status with my 15-20 fps in game” and even after going below any minimum video settings if it was possible💀💀
Will definitely be looking into this a bit closer. Thank you for the video!
Thanks for sharing! Seems easy and simple to follow! Can't wait to create great content using this!
Love this whole preview/tutorial, I'm personally not very familiar with practical applications of AI so watching stuff like this is really neat!
It's amazing what AI can do for video generation with AORUS 17X. Thanks for the fantastic video explaining all of it. Keep 'em coming!
Pretty cool that you can generate clips like that on your local system now, though the technology clearly has plenty of limitations and there's lots of room for future improvement.
I think this new Ai video generation feature is a true gamechanger, and just goes to show how far we've come in regards to technology.
It's amazing how far this tech has come in such a small time
i love the whole preview honestly this was very well done and it is free which is great!
It's a whole new world being able to make videos with ai. I might live through something absolutely incredible.
Going to be interesting to follow the effects of AI on everything.
Thanks for the quick 'how-to' on the install procedure.
Thank you for sharing. It seems easy and simple to follow! Can't wait to create content using this.
feel like ai video generation is gonna take over the scene of video editing by the upcmoing 5-10 years & be the normal standards from now on
Thanks for this video! The setup is surprisingly easy and it's awesome to see the capabilities of AI!
It's insane this stuff is running locally. I can't wait until a truly _open_ competitor to OpenAI's Sora makes waves.
Audio only playing on left channel :(
okay that is super weird.. perhaps I can fix this
I attempted to trim this video to see if UA-cam's video editor rerenders it in mono. it seems that there was an issue with the mono audio recording being converted to stereo on accident, my apologies.
@@MattVidPro I'm praying for a K.I. tool to come out soon that can perfect audio stereo distribution - Little joke, you're the best :) Wink 😉, this comment was crafted with a little help from ChatGPT
Sounds fine from my phone
Not a big deal, be thankful you aren’t deaf. She’ll be right mate
I only heard the audio on the left side,why ??
By the way loved the video about AI❤❤.
I've been using open-source AI stuff, like XTTS, Fooocus, and this one. Yes, it can be a hassle installing open-source things and waiting for it to be installed. But if you have a good PC you can do many things that other are paying for, and that is a win in my book. Plant the seeds of struggle, reap the harvest of ease.
Having AI being this freely accessible will be both dangerous and positive. It lowers the barrier of needing expertise but with that comes both good people being able to make good things and bad people making bad things.
Great tutorial. Lots of great stuff I hadn't heard about in Pinokio to experiment with. 👍
Local AI is quite nice. NVIDIA has been making great strides in helping make that more commonplace. Great tutorial, thanks!
Really informative, AI video generation seems so cool, can't wait to try it
Thanks for the video!!!
Pretty easy tutorial Matt!!
IA is here and It going to stay!!!
Thanks again!!☺️👌🏻
Is the audio only coming through the left channel for anyone else?
No, but it's mostly on the left.
Big problem with this is it doesn't tell you the size of the installs. Not good if you're trying to manage space, some of the installs are quite large and you have no idea how much of a dent they'll make until you've installed them.
Thanks for sharing! I find your channel to be informative and easy to watch.
Like the cat's crazy eyes show, the tech still has a long way to go. But it is nice to see even the basic forms already usable with consumer hardware.
It is amazing how much AI capability they can put into something like the AORUS 17X as it is comparatively small to a desktop PC.
Hey matt, Thanks for sharing with simple instructions! I appreciate it!
Thank you for the detailed video, I might try this!
Simple, straight forward video on AI. I already have a few ideas in mind.
The implications of AI are tantalizing and concerning all at once. Its certainly an exciting time to be alive
Looks promising... I am really excited to see what it will deliver in next 4 or 5 years!
I love that your doing videos with pinokio.
Awesome video! Might give this a shot on my days off and see what kind of stuff I can get going
Great video. Thanks. Much easier than I thought it would be. Going to have fun with this...
I'm so late at discovering pinokio, absolutely amazing thx for that :')
Your voice is only on the left channel. I thought my headphones are broken, but it is actually your video
Great demo of AI video generation on AORUS 17X by MattVidPro AI. Exciting to see AI's potential in content creation!
Wow, these tips would save a TON of time during editing - I would love to try some of the AI video generation programs you showed on the AORUS 17X !!!
How can you make it longer?
Really cool stuff! Seems pretty straightforward.
absolutely wild - images are one thing (and still very impressive), but videos? crazy!
Great tutorial Matt! Made it easy even for this Luddite! A+
Wow! The AI technology is improving so fast!
I tried to install pinokio on a 2021 M1 Mac…. does not work.
Can anybody help?
I hope to try this technology, I have only used Windows 7 and Chromebook computers, I have never even seen the later and evolved versions of computers and PCs.
Liked for the sponsor segment, relevant, don't need it, but if I did it looks like a good choice.
I don’t accept most sponsorships only the best ones make it thru! Really happy they were able to sponsor a tutorial like this that benefits everyone
Thanks for the video. In my opinion this is the new era to the technology.
Pretty sweet software to tinker with. Thanks for the guide!
Nice and very easy to follow even for someone who has nearly no experience with computers so far. I would have preferred to see directly what the hardware requirements are, not just looking at task manager and hearing chances are pretty good if...
Keep up the good work!
This is actually really cool, thank you!
Just watched the latest SORA AI generated videos... mindblown that we are reaching such a level this fast
on linux more than half will not work. if you install them with git and conda they work great. i think it can't handle the different envs good.
Looks really cool and I love that it's newb friendly.
Pinokio sometimes works and even the devs have said it's better to run on Mac and windows users are waiting their time
Hey guys! If you get the error “streamlit’ is not recognized as an internal or external command, operable program or batch file” perhaps the streamlit module is not integrated into python, pip install streamlit.
Awesome video, very informative thank you for your wisdom and teaching!
AI is blowing up! It's truly amazing what AI can do and what you can do with AI and this is only the beginning!
Not really for me, but I can difinitely see people making some really cool stuff with this.
Is this a newer method of doing this locally?
yes
Thanks for sharing. I learned something new today!!!
Seems to work fairly well for video generation. Looking forward to AI being more integrated into video games, particularly NPC interactions 🙂
Thanks for your hard work and the great and informative video!
always fails when attempting to load models with Connection Error Connection failed with error 0
AVAST! keeps deleting files seconds before the end of the install. How can this be avoided?
Would the new Macbook Pro M4 enough to run this?
Thanks for the video. Very informative and interesting setup
Really informative, thanks for the detailed explanation. I can't wait to start with AI.
Stable video diffusion is pretty nifty
I found this video very informative, especially compared to most tutorial videos.
Great info. Can't wait to try it out.
It's kind of disappointing that it has such a limited time generation. The Forge version of SVD allows you to adjust the clip length. Maybe I'm missing something but I couldn't get anything better than 1 second.
Installation of ffmpeg failed at first, I had to update my conda install, after that it worked.
I already had an old version of Pinokio, that yielded so many errors, that I gave up back then.
Maybe it was the old conda version all along 🤔
Thanks Matt 👍
I'm intrigued by its potential for AI video generation. The powerful specs, including the GeForce RTX 4090 GPU and 13th Gen Intel i9 processor, sound amazing.
I'm not sure what to think. I finally feel like I'm in "the future".
Cool how easy it is do so much! Reminds me of the early days of music generation software
Very informative and interesting. Much appreciated.
thank you so much for making this video bro!!!
I only have 6 GB dedicated GPU memory. Good night, folks!
I don't even have gpu lol
Good night
i run svd on a 6gb vram Nvidia 2060 laptop through Comfyui. i havent tried or needed pinokio
I have 24gb. If u send me sum bitcoins I can share.
Thanks for the info and keep up the good work. Nice cat 😻
Matt! Thanks so much this is awesoooome!
Damn, that's quite speed up for many things for creators and animators. Quick draft videos to polish, or just visualise thoughs💪🏽
Nice simple tutorial thanks for the video. AI is going to have a greater impact on all aspects of our lives as it improves and we find new ways to utilize it.
Has anyone received this error?
"LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on."
I can see that it exists locally. But it isn't able to access it.
Matt's stable video diffusion licensing doesn't allow for UA-cam videos and commercial use, am I right?
Nobody can own your images/videos and how would they even know
@Gafferman it probably puts some meta data in but you could remove it. But if you really making content you'd probably just use a paid service eventually.
Thanks for this great video! I enjoy your content.
i got virus alerts from some of the pinokio installs.
Any chance to run this on a VM ?
Another issue is that it installs versions of programs that can create issues with other AI tools.
I had a play on this for a few hours, was getting poor results. I found the clips too short and if u extended them the image would fall apart.
Maybe some1 out there does a deep dive into getting great results.
It's still early days for AI video, rn it's slow and results are super inconsistent.
to be fair, Stable diffusion is known for crap base models. The only reason it's so popular and widely known is the community that improves and build upon it. It wouldn't be insane to see in 1 year, 'models' for specific use cases so that way depending on what you want to do, you'll get better results based on that, (anime,people,animals,objects,backgrounds,etc) then you combine all models into one similar to SD image models. And probably the end result would be even better stuff than what the paid services would offer.
@@harnageaa seems like a hella lot of sliders and parameters.
Does my head in all that, you need effective simple solutions.
If your trying to make money you cant keep hoping and praying for tech and then theres learning and finding great reusable settings.
As it stands comfy ui & sd will never take off because no one wants to deal with all the $hit that comes along with it.
Eg : imagine how much wiring a record studio has. All the mics for drums, guitar, bass, keyboards, vox. They then go to amps, compressors, fx equipment... then into a 32 channel mixing desk ...
I own a DAW (digital audio workstation).
I simply record my audio & drop in fx.
These DAWs sell in their millions. Same with photoshop, blender etc.
The premium companies like midjourney will always be at forefront because they look to improve and simplify to make customers happy.
The only reason I am praising "open source models", is not because the models and the company offers an easy solution. I will give a direct example with image generation. After SD came out, there are now a TON of websites which took those "SD models" made themselves custom, and created AI PRODUCTS. Same for LLMs, they took the "free open source" models, tweaked, and made PRODUCTS.
What I care is not the RAW open source interface, but community going in, creating stuff, then others creating products with the results of the community.
Comfy ui is worst ui , in my opinion. It might have appeal for others but not for me. I'm more of a rectangle and sliders type of guy with tabs in it.
@@armondtanz
Motion bucket is amount of motion in video 255 is max
Interesting video tutorial. Watched and saved for future
This is great, i hope can run this without problems in my working laptop
I followed the install instructions but I don't get the links that take u to the stable video ui...
Are you sure the "T" field in the left-hand sidebar doesn't stand for "Time"? It makes sense to me that if you set T=14 and FPS=14, you get 1 second of output (14 frames at 14fps).
I think so too, I changed to 128 and it crashed, I changed back to 20 and it is still working
Thanks for the video! I know nothing about AI generation so thanks for the info
Honestly I'm a bit worried at how well some AI generated videos look already... It probably won't take long before they'll become indistinguishable from real videos 😨
Nice and easy to understand explanation!