2 ай бұрын
Пікірлер
@3r3n_5 3 сағат бұрын
Abi ne kasar taßaklí adamsın
@thesytx 14 сағат бұрын
Hocam 2 gündür sorunsuz kullanıyordum (davet bağlantıları ve profiller yüklenmemesi hariç) bugün artık mesajlar yüklenmemeye başladı. Sizde de böyle bir sorun var mı? Nasıl çözebilirim?
@SECourses 13 сағат бұрын
geçiçi olabilir bende şu anda sıkıntı yok. profiller yüklenmeme sebebi discord vpn ip adresini banlamış sanırım. tarayıcıdan baktım profil güncellemeye çalışınca too many request hatası veriyor - ip ile alakalı
@thesytx 13 сағат бұрын
@@SECourses hocam bende şimdi kontrol ettim tarayıcıdan, mesaj kanallarında "Too Many Requests" hatası alıyorum. Teşekkürler hızlı yanıtınız için.
@SECourses 13 сағат бұрын
@@thesytx aynen maalesef vpn ip si ile alakalı. mutlaka discord bunu gidermeli
@canopus1944 18 сағат бұрын
allow olarak bende çalıştı, tek sorun ben warp açında discordda mesajlar yüklenmiyor, sadece seste olanları görebiliyorum. En iyi ülke, en iyi 100 yılı. Eline sağlık kral, warp yüzünden oyunlara giremiyordum, sonunda girebiliom.
@SECourses 13 сағат бұрын
rica ederim
@TolgaKemal07 23 сағат бұрын
Hocam merhabalar 1 ay öncesine kadar tüm adımları izleyerek uyguladım discordu sorunsuz sıkıntısız bir şekilde kulllanabiliyorum ancak profil güncellerken veya bir discord sunucusuna katılmaya çalışırken problem oluyor ve herhalde bunlar içinde farklı adresler girmemiz gerekiyormuş eğer biliyorsanız benimle paylaşabilirseniz çok sevinirim
@SECourses 14 сағат бұрын
sanıyorum discord vpn den profil güncellemeyi engellemiş durumda. ip adresi bu ekli ama hata veriyor : 162.159.128.233
@TolgaKemal07 6 сағат бұрын
@@SECourses Evet her türlü VPN de engelliyor galiba şuan GoodbyeDPI türkiye özel versiyonunu kullanıyorum windows DNS ayarlarını yapıp yazılımı çalıştırdıktan sonra profilimi güncelleyebildim ve discord davetleri çalışır hale geldi şuan pek internetten yediğini ve arkada açık olmasının bir zararını görmediğimden dolayı dünden beridir kullanmaktayım ama WARP bilgisayarımda yüklü duruyor ne olur ne olmaz diye son zamanlar split tunel ayarı yapmaya çalışan arkadaşım başarısız oldu maillerine kod gelmediği için umarım en kısa sürede bu problemde kalkar ortadan cevabınız için teşekkür ediyorum ve önerim dilerseniz GoodbyeDPI hakkında video çekebilirsiniz türkiye özel versiyonunda superonline için bir çalıştırma seçeneği var DNS ayarını yapıp çalıştırıp yazılımı discorda rahatlıkla giriş yapabiliyoruz
@SECourses 55 минут бұрын
@@TolgaKemal07 bugün bende de warp vpn'i discord engelledi. az önce senelik kaspersky aldım mecburi çalışıyor. 200 TL. ama şikayet yazdım yorum yapıp paylaşırsan sevinirim : www.linkedin.com/feed/update/urn:li:activity:7266413332750901248/
@yjf-eb1gz 2 күн бұрын
Hello! I encounter a problem, could you tell me how to fix it? Thank you very much. when I run the GUI on a cloud computing platform. when I load the config file, the parameters and args remain the same. I can save the parameters to a new config file, but can not load any.
@SECourses 2 күн бұрын
it could be due to giving path inaccurately. on runpod it starts with /workspace
@SharpUpdate 2 күн бұрын
Does mimic pc have any lipsync software. Any software capable of lip syncing a video just like wav2lip
@SECourses Күн бұрын
they might have or you can use workflows. their discord : discord.gg/CdPbqFHJAr
@barackdrinksWD40 2 күн бұрын
it's not a "shocking truth" tesla has been transparent about this
@SECourses 2 күн бұрын
i didnt know until i saw this
@jyotibiswas-l1j 3 күн бұрын
I can't make high resolution (1080*1920) human pictures in the least invoke 5.0 people's pictures are showing double, high-resolution option are not show
@SECourses 3 күн бұрын
they have a great discord did you ask there?
@froxy61 3 күн бұрын
abi adamlar mail göndermiyo yani nası yapcaz bunu bide takım adım benim başka ama secourses yazıyom giriyo acaba ondanmı girmiyo (kendi takım adımı girince olmuyor)
@SECourses 3 күн бұрын
kendi takım adınla girmen gerekiyor. mail işi aynen dediğin gibi tekrar tekrar denemen gerekiyor malesef. by pass dene, email dene, allow all dene :D
@velRic 3 күн бұрын
why are you use only kohya to train the model? Kohya is a shit compare to onetrainer. the quality of UI in Kohya is a crap. Onetrainer is much better, very intuitive for beginners and give the same performance. I gave up when i started to read a documentation of Kohya. It is probably created only for asian users. Waste of time, especially once you don't have a super GPU
@SECourses 3 күн бұрын
OneTrainer coming from back for FLUX. Kohya has been over OneTrainer for a long time but recently OneTrainer catching up. You can see my onetrainer tutorial for SDXL here : kzbin.info/www/bejne/ZqWYnWl5hZ6YeMk hopefully i will make OneTrainer tutorial too
@FatalistCatalist 4 күн бұрын
Thank you for your video. Is it possible to use my desktop GPU remotely from my laptop on an external network?
@SECourses 3 күн бұрын
yes you can use with gradio live share or starting swarmui with cloudflared. checkout my main swarmui tutorial source link
@MacherTV 4 күн бұрын
Can you please make a tutorial on how to finetune mochi1 and train a lora or dreambooth for it? That would be amazing!
@SECourses 3 күн бұрын
i dont plan that at the moment - i dont know either :)
@zachtaurosa887 4 күн бұрын
what shoud I do if A6000 isnt available?
@SECourses 4 күн бұрын
A6000 alt config. if that is not also available you can go with anything you see you can afford. i am gonna let them know to add more
@zachtaurosa887 4 күн бұрын
@SECourses awesome thanks for yhe reply. looking forward to this!
@SECourses 3 күн бұрын
@@zachtaurosa887 i did forward your message lets see :D
@maxw8015 4 күн бұрын
@secourses i was wondering if you could do a tutorial on how to update automatica 1111 from version 1.6 on runpo to version 1.10 (latest) thank you!
@SECourses 4 күн бұрын
i have latest version runpod installer here : www.patreon.com/posts/86438018
@invaliddejavu 4 күн бұрын
maile kod gelmiyor:(
@SECourses 4 күн бұрын
aynen mail sistemleri çok bozuk. her ihtimali denemen lazım tek tek ve tekrar tekrar. adamlara raporladım ama sallamıyorlar. ama 1 kere çalıştırdın mı sürekli çalışıyor problem yok 7/24
@lorahester7 5 күн бұрын
Where can I see a proper dataset of images for training a flux Lora? How many portraits? How many full body shots, different emotions etc?
@SECourses 5 күн бұрын
here example dataset published the more is better : civitai.com/models/911087/dwayne-johnson-aka-the-rock-flux-dev-fine-tuning-dreambooth-model-for-educational-and-research-purposes-dwayne-johnson-aka-the-rock-flux-dev-lora-model-for-educational-and-research-purposes-full-tutorial Complete Dataset, Training Config Json Files and Testing Prompts : www.patreon.com/posts/114972274
@harrybarden5438 5 күн бұрын
Why are you inpainting the result images? Surely you want to test/evaluate the raw images directly from trained model?
@SECourses 5 күн бұрын
well the model cant even generate distant shots for random person so it is best to inpaint . on some closer prompts i dont inpaint i decide according to each prompt.
@user-zi6rz4op5l 6 күн бұрын
So you are the config peddler of open source project. Shameless grifter 💩💩
@jk7482 5 күн бұрын
Reddit scammer 😂😂😂
@jshnewman 6 күн бұрын
Thank you! We need more people like you in the world. I hope to follow in your footsteps one day
@SECourses 6 күн бұрын
thank you so much for the comment
@Blenderanswer-xk9vo 6 күн бұрын
Thank you so much for the video!
@SECourses 6 күн бұрын
you are welcome
@bentontramell 6 күн бұрын
Tip - stay away from political figures. Celebs/athletes?
@SECourses 6 күн бұрын
sure so you recommend Celebs right?
@wanderoffinc 6 күн бұрын
Jiga bytes bro .. Jiga bytes. good vid
@SECourses 6 күн бұрын
ye i still cant tell it accurately haha :D
@pavelratov-l4d 7 күн бұрын
automatic1111 not generated after training NansException: A tensor with NaNs was produced in Unet. This could be either because there's not enough precision to represent the picture, or because your video card does not support half type. Try setting the "Upcast cross attention layer to float32" option in Settings > Stable Diffusion or using the --no-half commandline argument to fix this. Use --disable-nan-check commandline argument to disable this check. if you apply the parameters a black picture is generated 3060 12 Gb
@SECourses 6 күн бұрын
try with swarmui and see if works : kzbin.info/www/bejne/Y6uyfJulhJ6Ua9U
@pavelratov-l4d 6 күн бұрын
@@SECourses in automatic1111, it began to generate after I set these parameters to 32 during training - Override U-Net Data Type: float32 Output Data Type: float32. the generation time has become longer, but now the automatic1111 accepts trained weights.
@SECourses 6 күн бұрын
@@pavelratov-l4d great
@AripSam-ls3vy 7 күн бұрын
Can you make a Colab pro notebook for easyanimate pls? Or explain how you can make one?
@SECourses 7 күн бұрын
you can use our runpod or massed compute installers
@Because_Reasons 7 күн бұрын
Do I need regularization images? I get a warning none found.
@SECourses 7 күн бұрын
for flux you dont need
@edhooks4605 7 күн бұрын
None of the options for installing Python for all users exist on my computer. I have tried running python install as individual user and as administrator and the boxes for all users are are grayed out in both cases. Any suggestions?
@SECourses 7 күн бұрын
your current account may not have enough privileges. it could be prevented by antivirus or your computer might be under infection. also after this i recommend this video : kzbin.info/www/bejne/eqPLhnukjsuqeZI
@geminimanvfx 7 күн бұрын
in _create_inference_session sess = C.InferenceSession(session_options, self._model_path, True, self._read_config_from_model) onnxruntime.capi.onnxruntime_pybind11_state.NoSuchFile: [ONNXRuntimeError] : 3 : NO_SUCHFILE : Load model from .\models\det_10g.onnx failed:Load model .\models\det_10g.onnx failed. File doesn't exist in _create_inference_session sess = C.InferenceSession(session_options, self._model_path, True, self._read_config_from_model) onnxruntime.capi.onnxruntime_pybind11_state.NoSuchFile: [ONNXRuntimeError] : 3 : NO_SUCHFILE : Load model from .\models\det_10g.onnx failed:Load model .\models\det_10g.onnx failed. File doesn't exist getting this error
@agrestige7390 7 күн бұрын
discord profillerini göstermekte sorun yaşıyor, bi tek bende mi bu sorun?
@SECourses 7 күн бұрын
bende sıkıntı yok daha fazla discord ip si bakıp ekleyebilirsin
@candymancls 8 күн бұрын
Thank you for this great Tutorial. I have a question.. I followed everything and I used 54 Pictures for the Lora training.. I have a rtx 4090 and choose Rank_3_T5_XXL_23500MB_11_35_Second_IT as the config file. I´ve started the Training 15 hours ago and now I´m on step 430/10800. So this will take forever.. Is this normal.. or should I choose a different config file or reduce the 200 Epoch? In your Video the steps go much faster.. it takes like 2 minutes or more for 1 step for me with this settings.
@SECourses 8 күн бұрын
that config requires huge VRAM. not suitable for rtx 4090 so you must be using shared VRAM. also today hopefully I will update configs and they will work faster. you should use rank 3 without T5 training. after todays update this T5 config may work fast as well stay tuned.
@candymancls 8 күн бұрын
@@SECourses Awesome, thanks for your reply
@SECourses 8 күн бұрын
@@candymancls you are welcome
@candymancls 8 күн бұрын
@@SECourses When I use the Rank 3 without T5 training it gives me this error: ValueError: T5XXL is trained, so cache_text_encoder_outputs cannot be used. I´m trying to make the Lora for the getphat FLUX Reality model from CivitAI which is a Dreambooth checkpoint I belive. Do you know what causes this error? I´ve turned Cache Text Encoder and Cache Text Encoder Outputs to Disk off and also Memory Efficient Save and the Training seems to start..
@SECourses 7 күн бұрын
@@candymancls yes it is true. by the way currently i am doing huge tests to update configs. so many cases to test. if you join discord i can show :D
@reapicus557 8 күн бұрын
Thank you very much for another great tutorial! Worked right away for me, and I appreciate the big list of example prompts to try out. I was hesitant to try another model, but after Civ added Mochi to their generator I knew I had to figure it out. Plus, I'm really happy that there is a 15 GB Mochi version that I can run locally. The results look much better than I was expecting, and generation times are not too bad. ৻( •̀ ᗜ •́ ৻)
@SECourses 8 күн бұрын
yep and hopefully something more amazing tutorial coming soon which supports image to video
@bayramtatar 8 күн бұрын
Merhaba Furkan bey, Videonun 23.04 dakikasında text encoderi nasıl değiştiriyoruz. biraz detay verebilirmisiniz. terminalde yükleme yapmıyo encederi. Manual olarak klasöremi ekliyoruz clip encoderi yoksa arayüzden mi indireceğiz ayrıca neden hangi dosyada isim değişikliği yapıyoruz.. teşekkür ederim.
@SECourses 8 күн бұрын
ismi değiştirmezsen otomatik yüklüyor. onun yüklediğine bak. sonra kendi indirdiğini onun yüklediği ile aynı isim yap ve eski dosyayı silip aynısı yap
@bayramtatar 8 күн бұрын
@@SECourses teşekkürler
@SECourses 8 күн бұрын
@@bayramtatar rica ederim
@sunnysihag7936 8 күн бұрын
Just stumbled upon your channel. Liked other videos didn't gone through but before starting the paid version on rupod i wanted to give a free try on kaggle, so opened this video now i can see you have made the kaggle post paid in your patreon community. Don't know I am disappointed or not.
@SECourses 8 күн бұрын
Kaggle banned the web apps. so we are keeping it hidden from public and it requires constant maintenance and update
@nylreia 9 күн бұрын
Hey, regarding Kohya setup, is that just the flux branch or did you add extra changes?
@SECourses 9 күн бұрын
this is just flux branch of kohya gui
@St4rg4zer_ 9 күн бұрын
iyi günler 8:00 gibi uygulama üstünden grup adıyla girmeye çalışıyorum ama sizin ki gibi bağlanmıyor.
@SECourses 8 күн бұрын
bazen çalışmayabiliyor tekrar denemek lazım. sonradan düzelen çalışan epey oldu
@Corgi_Leonidas 8 күн бұрын
@@SECourses ne kadar süreyle? aynı adımdayım ve ilerleme kat edemiyorum
@SECourses 8 күн бұрын
@@Corgi_Leonidas bilemiyorum ama sonradan çözüp çalıştıran çok kişi oldu. belki hatalı yaptığın kısımda olabilir. bende şu anda sorunsuz çalışıyor
@Corgi_Leonidas 8 күн бұрын
@@SECourses sonradan çalıştıranların da aldıkları hata Page isn't working, redirected you too many times mı acaba?
@SECourses 8 күн бұрын
@@Corgi_Leonidas aynen. email dene, allow all dene birde bunu dene: I found the solution for people getting ERR_TOO_MANY_REDIRECTS after trying to connect through Zero Trust, so if in Device Enrollment Rules bypass is not working (which was my case) I tried Allow and put my email. Then when I login with Zero Trust it asked for the email and sent a code to enter. After it worked, now some people might be confused (like me) because when I went on advanced in Zero Trust preferences it still said "Exclude Split Tunnels" and to change that you have to follow these steps: In Zero Trust, go to Settings > WARP Client. Under Device settings, locate the device profile you would like to modify and select Configure. Scroll down to Split Tunnels. (Optional) To view your existing Split Tunnel configuration, select Manage. You will see a list of the IPs and domains Cloudflare Zero Trust excludes or includes, depending on the mode you have selected. We recommend making a copy of your Split Tunnel entries, as they will revert to the default upon switching modes. Under Split Tunnels, choose the mode: Include IPs and Domains After that we can follow the video to the end
@nextlevelbrosagency 9 күн бұрын
This has been the most useful video since I began my AI journey. My PC was a ad-hoc mess of versions of a million things installed one by one as needed to get things to work and I was always afraid to mess up previously working setups, with all the requirements and conflicts and after a point, contradictory requirments for different tools. Now I understand what I am doing when it comes to versions for Python, CUDA etc. The Environment Variables explanation was also gold. Thank you bro, you're the best!
@SECourses 9 күн бұрын
thanks for comment. ye i know these are the gold things that no one else explains. i hard way learnt :)
@nextlevelbrosagency 9 күн бұрын
@@SECourses That's the truth, people show fun AI tools but never actually help when people run into issues with installation and error. You put in the time to learn it all and to share it all, I am always amazed. BTW a heads up, on your github Patreon Posts Index, the EasyAnimate link (the latest Exclusive post) is liking to the 'Free Kaggle Account' post (the 2nd to latest post).
@AndrewIokeek 9 күн бұрын
I've had alot of enthusiasm in learning SE as a self taught, since I was in high school , but wasted my three years of my life being stuck on the beginning mindset couldn't make a move because I thought I needed quite alot of support and knowledge in Order to understand and concentrate on learning, hence it's really satisfying that today I came across this cause and it's really clear and motivating for push forward. Time to get my schedules in order and in a better way this time..... I wasted some time trying to learn Python but failed because the was no fundamental structure and derection of what am Aimed... Hope this will get to be handy...school is expensive after high school ....trying to have more courage and take an advantage as a self taught...this videos gonna help me manage my schedule between work and education, wish there could be a group of people like me here in the section so we can connect and try to learn together form a little community of enthusiasts thanks to SE✅
@Maria-o2e2t 9 күн бұрын
The yolo classes at 23:30, you checked them and made a folder, but what's the point. What if you don't do it?
@SECourses 9 күн бұрын
it is for auto masking face and inpainting. not mandatory but helps significantly
@Maria-o2e2t 9 күн бұрын
@@SECourses thank you, then I definitely need that, face is the most important part of my pics :)
@SECourses 9 күн бұрын
@@Maria-o2e2t 100% use it
@lucifer9814 9 күн бұрын
WHY IN THE FUCKING HELL ARE YOUR TUTORIALS SO BITCHING LONG ? UNFORTUNATELY THERE'S NO OTHER TUTORIAL FOR THE INSTALLATION PROCESS BUT I CAN'T SIT HERE WATCHING YOU RAMBLE ON FOR 2 HOURS.
@vida35 9 күн бұрын
Emeğine sağlık hocam tam aradığım çözüm
@SECourses 9 күн бұрын
Rica ederim teşekkürler yorum için
@KuiDong-b2u 10 күн бұрын
Blogger, I was wondering if this model could use API calls to generate videos ?
@SECourses 9 күн бұрын
yes swarmui has api you can use it
@markschrader7869 10 күн бұрын
Do you know how to get Automatic1111 to recognize FLUX DEV GGUF file versions? And are you aware that Nvidia recently announced the Flux models now support the NVIDIA TensorRT software development kit, which improves their performance up to 20%?
@SECourses 9 күн бұрын
automatic1111 doesnt support FLUX, you can use forge web ui. swarmui has tensorrt mode i plan to make a tutorial for it
@markschrader7869 9 күн бұрын
Thank you for your quick reply, have you been able to get the TensorRT extension to work with Forge Webui? I have not had any previous success with installing TensorRT with Forge. I did with your Automatic1111 tutorial. Which reliable Webui would you recommend for Windows 11 & Flux with TensorRT? I'll be using with RTX 3070 8gb vram & FLUX1_dev_Q5_1.GGUT OR FLUX1_dev_Q4_1.GGUT
@SECourses 8 күн бұрын
@@markschrader7869 sadly i dont know that atm. you should try both but it might not be supporting GGUF
@848 10 күн бұрын
XN convert is very useful for the resizing, you can batch resize and crop all the images in under a second
@SECourses 9 күн бұрын
nice thanks for tips
@DawidSikora 11 күн бұрын
After doing these steps in disk management I see the correct total size of 1860 GB in C, however in file Explorer I see 439 GB... Any idea what happened? Edit: going into diskpart->select volume->extend filesystem has fixed the issue Thanks for the video!
@SECourses 11 күн бұрын
awesome thanks for the info
@johnnyapbeats 11 күн бұрын
Why so low frame rate?
@SECourses 11 күн бұрын
it supports 8 FPS but up to 1024px HD resolution
@bassiedappie635 11 күн бұрын
Hi AI noob here. I watched a lot of tutorials on Lora's from different creators. This is the first tutorial I watched on Fine-tuning. I noticed you didn't provide captions for the dataset when you started training the AI. What is the reason behind this? Is this not necessary on dreambooth or not beneficial? As far as I understand from my very limited experience training Lora it helps when you provide the captions so I am curious.
@SECourses 11 күн бұрын
it is not beneficial for FLUX, I have tested it so many times now. If your dataset is consistent, e.g. only a single subject, object, style etc, captions only reduces quality and likeliness. so we only use rare token + class token and it is read from the folder names
@sathvikpasunoori9560 11 күн бұрын
make a video on installing locally of easyanimate v5
@SECourses 11 күн бұрын
i published the installers but will make a video too : www.patreon.com/posts/easyanimate-1-to-115888558
@DaVillen 5 күн бұрын
@@SECourses of course you just want money.
@SECourses 5 күн бұрын
@@DaVillen If you are working for your boss for free please start working for me. ty
@DaVillen 4 күн бұрын
@@SECourses Only grifters want to make money from a hobby and ruin for everyone over greed.
@DaVillen 4 күн бұрын
@@SECourses You will be banned from every subreddit, the community fucking hates you!!
@Nathan-ib2rs 11 күн бұрын
So it basically is runway with lora support?
@SECourses 11 күн бұрын
yes but not that good yet