- Видео 98
- Просмотров 65 338
Devs Kingdom
Добавлен 7 мар 2022
Jupyter, Kaggle and Colab is all you need to learn and get real hands-on experience with AI and most Technologies. Either as a starter or professional, you will benefit a lot from it. This Channel provides Tons of Tutorials to show you how, and meanwhile Make Tech & Open Source Communities more accessible. Welcome and Kindly stay tuned for every video to Max your value.
This Channel is Formerly (Jacob Dev)
General Inquiry:
support@productdeploy.com
For any legal or copyright issues, credit & inquiries, please contact me on -
legal@productdeploy.com
This Channel is Formerly (Jacob Dev)
General Inquiry:
support@productdeploy.com
For any legal or copyright issues, credit & inquiries, please contact me on -
legal@productdeploy.com
Browser Use + Ollama + QWen 2.5: Automate Websites Access for AI Agents (Local Install)
This video demos how to build Browser-Use AI Agents with Ollama to access websites on Kaggle as an automation process (rather than the official readme method which runs browser-use on a desktop environment)
Kaggle notebooks:
Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a few reasons. 1. it works while the video is recorded, However it does not guarantee to work at a later date as tech communities make code changes all the time. Please follow the tutorial and create your own version of it if needed. 2. If you have any questions or need help, please join the discord serv...
Kaggle notebooks:
Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a few reasons. 1. it works while the video is recorded, However it does not guarantee to work at a later date as tech communities make code changes all the time. Please follow the tutorial and create your own version of it if needed. 2. If you have any questions or need help, please join the discord serv...
Просмотров: 112
Видео
Roo-Cline (Cline 3.1 Fork) + Ollama + Coder Server + Visual Studio Web: Best Open Stack AI Coder Yet
Просмотров 2764 часа назад
This video demos how to setup Roo-Cline (Best Cline Fork) with Ollama and Visual Studio Web from "Coder Server" on Kaggle Free GPU Code Server Visual Studio Web on Kaggle might be even better than Google IDX. Kaggle notebooks: Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a ...
MoonDream2: Best Open Source Tiny Vision Language Model Local Install and Test
Просмотров 1297 часов назад
Kaggle notebooks: Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a few reasons. 1. it works while the video is recorded, However it does not guarantee to work at a later date as tech communities make code changes all the time. Please follow the tutorial and create your own ve...
Promptimal + Ollama + Google IDX: Best Open Source Prompt Optimizer
Просмотров 2377 часов назад
this video demos how to setup Open Source Prompt Optimizer - Promptimal with Ollama on Google IDX Kaggle notebooks: Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a few reasons. 1. it works while the video is recorded, However it does not guarantee to work at a later date as ...
Llama Coder + Ollama + Qwen2.5 + Postgresql: The Best Open Stack Web AI Coder Local Install and Test
Просмотров 1,2 тыс.9 часов назад
This video demos how to build Llama Coder with Ollama and Postgresql locally without Together AI Step by Step Kaggle notebooks: Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a few reasons. 1. it works while the video is recorded, However it does not guarantee to work at a la...
Dolphin 3.0 + Ollama + OpenWebUI: Ultimate General Purpose Steerable LLM Install and Test Locally
Просмотров 18314 часов назад
Kaggle notebooks: Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a few reasons. 1. it works while the video is recorded, However it does not guarantee to work at a later date as tech communities make code changes all the time. Please follow the tutorial and create your own ve...
Cline 3.1 + Google IDX + Ollama + Qwen 2.5 (Upgrade):Best Open Stack AI Coder is Getting Even Better
Просмотров 89814 часов назад
this video demos the latest Cline 3.1 changes (Checkpoints, Restore and Compare) with Google IDX Ollama Qwen 2.5 Kaggle notebooks: Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a few reasons. 1. it works while the video is recorded, However it does not guarantee to work at a...
Qwen SmallThinker-3B: Reasoning Like GPT-4O with Ollama and OpenWebUI
Просмотров 15616 часов назад
Kaggle notebooks: Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a few reasons. 1. it works while the video is recorded, However it does not guarantee to work at a later date as tech communities make code changes all the time. Please follow the tutorial and create your own ve...
Agentarium: AI Suite Powered Interactive Agents Install Locally
Просмотров 26021 час назад
This video demos how to install and Test Agentarium Agents with Ollama, Groq and Huggingface APIs on Kaggle GPU Kaggle notebooks: Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a few reasons. 1. it works while the video is recorded, However it does not guarantee to work at a ...
Microsoft Trellis (Upgrade): Best Open Source Image to High Quality and Versatile 3D Generation
Просмотров 59021 час назад
This video demos how to build Microsoft Trellis Image to High Quality 3D on Kaggle Free GPU Step by Step including "How to manage multiple CUDA versions locally" Trellis Features: High Quality: It produces diverse 3D assets at high quality with intricate shape and texture details. Versatility: It takes text or image prompts and can generate various final 3D representations including but not lim...
SmolAgents: Best Free Simple Library to Run Powerful Agents in a Few Lines of Code - Local Install
Просмотров 209День назад
SmolAgent is a Open source library that enables you to run powerful agents in a few lines of code. It offers Simplicity, First-class support for Code Agents and Hub integrations This video demos how to build it on Kaggle for Free Kaggle notebooks: Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a...
SmolAgents: Best Open Source Library to Run Powerful Agents in a Few Lines of Code
Просмотров 182День назад
this video demos how to run SmolAgents with Ollama and Groq on Google Colab 100% Kaggle notebooks: Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a few reasons. 1. it works while the video is recorded, However it does not guarantee to work at a later date as tech communities ...
CosyVoice 2.0: Best Open Source Full-Stack Multi-lingual Large Voice Generation Model
Просмотров 337День назад
Kaggle notebooks: Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a few reasons. 1. it works while the video is recorded, However it does not guarantee to work at a later date as tech communities make code changes all the time. Please follow the tutorial and create your own ve...
DeepSeek V3 Free API + Bolt.DIY: New OttoDev AI Coder Install Locally Step by Step
Просмотров 1,2 тыс.День назад
Bolt.DIY is formerly ottoDev which is the best Open Source version of Bolt.new from Stackblitz. in this video, it demos how to use Bolt.DIY with DeepSeek V3 for free Step by Step Kaggle notebooks: Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a few reasons. 1. it works while...
Gemini Coder + Google IDX: Best Open Stack to Test Google LLM Locally
Просмотров 44914 дней назад
Kaggle notebooks: Notebook is only created for demonstration and serve as a guidance for those who were interested using similar methods to build projects. It is NOT a free giveaway for a few reasons. 1. it works while the video is recorded, However it does not guarantee to work at a later date as tech communities make code changes all the time. Please follow the tutorial and create your own ve...
DeepSeek V3 Free API + Google IDX + Cline 3.0 + Aider: Best Open Source Huge Model Test Locally
Просмотров 4 тыс.14 дней назад
DeepSeek V3 Free API Google IDX Cline 3.0 Aider: Best Open Source Huge Model Test Locally
Hunyuan Video GGUF (Upgrade): Best Open Source Low VRAM HD UpScale Pipeline Local Install Tutorial
Просмотров 1,1 тыс.14 дней назад
Hunyuan Video GGUF (Upgrade): Best Open Source Low VRAM HD UpScale Pipeline Local Install Tutorial
ScrapeGraphAI: Best Open Source LLM Scraping Pipeline Framework
Просмотров 18214 дней назад
ScrapeGraphAI: Best Open Source LLM Scraping Pipeline Framework
Qwen QVQ 72B: Best Open Source Visual Reasoning LLM Beats O1 and Claude Sonnet 3.5
Просмотров 22114 дней назад
Qwen QVQ 72B: Best Open Source Visual Reasoning LLM Beats O1 and Claude Sonnet 3.5
DeepSeek-V3: Best Open Source Huge Model Beats Claude Sonnet 3.5 in Coding (Fully Tested)
Просмотров 2,4 тыс.14 дней назад
DeepSeek-V3: Best Open Source Huge Model Beats Claude Sonnet 3.5 in Coding (Fully Tested)
Aider + Google IDX + Ollama + Qwen 2.5 : Best Open Stack Inline Text to Code
Просмотров 18414 дней назад
Aider Google IDX Ollama Qwen 2.5 : Best Open Stack Inline Text to Code
LTXV GGUF (0.9.1 Upgrade): Best Low VRAM Model for Super Fast Realtime HD Video Generation
Просмотров 55014 дней назад
LTXV GGUF (0.9.1 Upgrade): Best Low VRAM Model for Super Fast Realtime HD Video Generation
GroqLabs Llama 3.3 AppGen: Super Fast App Generation Local Install Step By Step
Просмотров 37421 день назад
GroqLabs Llama 3.3 AppGen: Super Fast App Generation Local Install Step By Step
Cline 3.0 + Google IDX + Ollama + Qwen 2.5: Best Open Stack AI Coder
Просмотров 4,3 тыс.21 день назад
Cline 3.0 Google IDX Ollama Qwen 2.5: Best Open Stack AI Coder
Virtual Try-On: Best ComfyUI Workflow with Flux.1 Fill GGUF + Redux + Turbo Alpha and CatVitOn Lora
Просмотров 37421 день назад
Virtual Try-On: Best ComfyUI Workflow with Flux.1 Fill GGUF Redux Turbo Alpha and CatVitOn Lora
Microsoft MarkItDown: Convert Files and Office Documents to Markdown (Local Install Step by Step)
Просмотров 1,3 тыс.21 день назад
Microsoft MarkItDown: Convert Files and Office Documents to Markdown (Local Install Step by Step)
Hunyuan Video GGUF: Best Open Source Low VRAM Model for Video Generation Local Install Tutorial
Просмотров 1,4 тыс.21 день назад
Hunyuan Video GGUF: Best Open Source Low VRAM Model for Video Generation Local Install Tutorial
OmniAudio-2.6B: Best Small and Efficient Audio Language Model from Nexa AI (Edge Device Support)
Просмотров 11521 день назад
OmniAudio-2.6B: Best Small and Efficient Audio Language Model from Nexa AI (Edge Device Support)
Microsoft Phi-4: Best Open Source Small Language Model (Fully Tested)
Просмотров 22121 день назад
Microsoft Phi-4: Best Open Source Small Language Model (Fully Tested)
LTX-Video: Real-time HD Video Generator (Low VRAM LTXV ComfyUI Local Install Tutorial)
Просмотров 39221 день назад
LTX-Video: Real-time HD Video Generator (Low VRAM LTXV ComfyUI Local Install Tutorial)
hi dont stop the music
hello, can you share the .clinerules file ? thanks
so helpful video
Did you go to the store and say to the salesman, "Give me the worst microphone you have"? Then you came home and swallowed it when you recorded this so we can hear every sound your mouth possibly make? (I check your new videos and you fix this issue but I send this comment anyway😘)
thanks for the comment, glad it worked
as a windows 11 pc user , I tried to open that Kaggle Notebook Mystery, but whatever I code into these "Cells" nothing happens. would be helpfull to at least show how you start in that Kaggle-notebook world, all the kaggle Tut for beginners on You tube are presented by Indian native Code freaks on speed.
⚠ when you ask it who is it? it tells you its actually CHATGPT model and not deepseek at all ⚠
there were discussion on this, seems due to training data, not the model
The complexity of the explanation is too great.
thanks for the comment
Which video is the one where you show how to setup openwebUI in Kaggle?
latest kaggle version ruclips.net/video/0jAhK3hlIbM/видео.html
older kaggle version with conda ruclips.net/video/lcnybgwqqBg/видео.html
That music :D From now on when I ask a question to any llm,it will play in my brain automatically. Thanks for the test!
I have 8gb vram how much will it take for 5 secodn vidoe ? Thank u in advance ❤
have not tried 8G yet, Kaggle has 16G VRAM which tested in the video. you can register for free and give it a try. Best,
i dont know how to start
search google idx and register
This is perhaps a trivial question.. but does context length matter a lot for the LLM model being used to power the agents? Can Agentarium still be effective in the case where the context is limited to 8K?
context length matters more to what the agent asks, rather than the effectiveness or capability of agents
Gpu minimal?
min VRAM not tested.
how can i set the /set parameter num_predicts <int> in this script?
create a modelfile, example github.com/ollama/ollama/blob/main/docs/modelfile.md#basic-modelfile
please also feel free to join the discord channel to discuss
Dolphin 3.0 has been released!
feel free to checkout this video ruclips.net/video/0jAhK3hlIbM/видео.html
Everyone is showing us videos of making apps with ai but now how to actually compile those apps into executable desktop apps, if you could do that you will be the first and that would be great
Trellis Code has been updated. here is the upgraded tutorial to Setup Trellis on Kaggle ruclips.net/video/U6QBvCeWpZI/видео.html
Thought this was going to be the solution, but got to hyperbolic and absolutely and totally lost. I feel you need to explain fully as you go as through and take the screen a bit slower with more emphasis on where the clicks go. Anyway, that's only me i guess I may be of the very few here, so thks for the try.
thanks for the comment
also feel free to join the discord channel if you have any questions.
@@DevsKingdom Have tried many times to use Discord but failed sir.
Can you please share the code?
@@sriniBw pls refer to notebook sharing claims in the video description, thanks
add the code file
@@CarToon-m1m7o pls refer to notebook sharing claims in the video description, thanks
Thanks for sharing such valuable information! A bit off-topic, but I wanted to ask: My OKX wallet holds some USDT, and I have the seed phrase. (alarm fetch churn bridge exercise tape speak race clerk couch crater letter). How can I transfer them to Binance?
thanks for the comment
Very nice. Thanks
This is great but I wish fine tuning was released for it. Im waiting for fine tuning to be released for thus and mask-gpt. The developer for this one said they wouid release a fine tuning implemention, but when asked when they didnt reply. That was 4 days ago. I dont know if Mask-GCT devs ever intend on releasing a fine tune model, they last said there were working on the next gen of mask-gct in october, but I saw some of the devs for that just released vevo a week ago so idk if they stopped mask-gct.
Hi, Im marco from scrapgraphai, I would really like to say thank you for the video you have done, keep believing in your projects and your dreams that sooner or later will come true. wish you the best, Marco
you are welcome !
wait whats with the name change ?
channel name
can I run it on a really potato laptop?
feel free to give it a try
Jacob to Kaggle Kingdom now is Devs Kingdom...
@@mallardlane8965 should be allset :)
cline error API Request Failed Unexpected non-whitespace character after JSON at position 113 (line 1 column 114) how to fix it ?
explained in discord channel
i have 4gb gpu nvdia 1650 will it work fine ??
that might be tough
Does the OpenRouter API provide the same performance as DeepSeek's?
didnt try openrouter.
@@DevsKingdom Alright! Thanks.
Thank you for the video mate. I set up the settings for Hyperbolic just as in your video and I am getting an error: API Request Failed Unexpected non-whitespace character after JSON at position 113 (line 1 column 114) Any idea what could it be? Thanks!
kindly join the discord channel to discuss
@@DevsKingdom Just did. Thank you!
same here
pinged u in discord
what i don't like about cline is millions of tokens for a simple html css webpage... waiting for planning, waiting for each request... even copy paste from chatgpt's website would be simpler 😅
lolz make sense
hey could you atleast make the notebook publically available on kaggle profile of your just for now it has become hard for debugging the errors im getting when trying to run it
please kindly check the notebooks sharing claims in the video description
@@DevsKingdom Where? There is no Notebook Url in Description
You’ve got some solid info here, but honestly, the delivery makes it hard to get through. The audio quality and unclear pronunciation make it frustrating to follow, especially for a technical topic. It feels like you’re rushing through key details without ensuring the audience can actually understand them. I’m here for the content, but it’s a struggle when the presentation gets in the way. A bit more polish on clarity and pacing would go a long way. Just being real-hope this helps!
thanks for the comment
audio is very quiet .. cant hear anything. its a long video - it would help if you can break it into sections: intro, some important points, conclusions
@@DeianGiNet thanks for the comment. this is an older tutorial. pls kindly checkout the latest videos in this channel and let me know if u get better quality.
I use roocline... is it better to use cline normal version?
thanks for the comment. have not tried roocline yet.
i try use it with roocline or cline, but i have this error: "Unexpected non-whitespace character after JSON at position 113 (line 1 column 114)"
pls kindly join the discord channel to discuss
Thanks for the information. Question. My cline keeps saying it reaches 64k content maximum. Do you have that problem? A little constructive criticism if you dont mind: speak louder and clearer.
api-docs.deepseek.com/quick_start/pricing
Good video
thank you for the tutorial , but a 11.1m tokens is too mush for a simple page unless you did some thing edited in the video !!
@@abdelouahabmotrani3831 no edit . maybe the tokens were also counted for other tests not showed in the video
Thanks for the videos! I just wanted to ask if anyone who has tried this method with Cline and VSCode is experiencing this strange issue: the AI responds to questions, but it replies based on another project that isn’t mine, and I have no idea where it’s getting it from!
Cline detects the files of the folder it is working on. maybe create a empty folder and set a rule in .clinerules to specify to add all the files to that folder
Nice tutorial! how much time does it take on the sampler node at 720*416 resolution for 3 sec clip? I am using fp8 version and it takes around 15 mins for a 3 sec clip on 434*240 resolution on P100 gpu. I've used gguf before but doesn't make any difference on time for a 3 sec clip
yes. 10 to 15mins make sense.
hi, thanks for your video, I wanted to know if Android or iOS apps can be created on idx google, I had seen with android studio, I applied but I don't think they accepted me, and then in any case would it be possible to export an app from idx?
idx android studio is on experimental, have not tried that .
yes i think they apparently don't accept anyone yet but is there a way to create and export an android or ios app?
@@djekwkdkwwdjekwkdkeoqo idx is based on nixOS, here is the android doc on nixOS nixos.org/manual/nixpkgs/unstable/#android also in the project, you should have a .idx folder, there should have a nix.dev file, you can add android dev tools cli packages.
I'm in the discord, i write to you
Thanks for the videos! Would you mind sharing details on the .clinerules file you use and any other best practices for better coding?
thanks for the comment. clinerules in the tutorials is very basic (1. write all the generated files to a specfic folder and 2. log changes for each update to a markdown file. e.g. changes.md). .clinerules can be setup for each project depends on the project details. it is like system prompt, there could be better templates online.
thanks boss
Thanks to you, I managed to get Groq_AppGen to work.
Hunyuan Video GGUF Upscale Pipeline with Low VRAM support on Kaggle ruclips.net/video/Z1fD6SJCj3Q/видео.html
You should have explained how you git clone using kaggle, you can't just skip important steps like that. Also for fucks sake, get a better mic. Absolutely useless tutorial, you didn't show any steps or provide the text to copy and paste to install.
git step on kaggle is at 1:20 . mic was fixed, feel free to check latest videos, thanks.
@@DevsKingdom Yes but for a new user you should provide step by step text so a user can follow it properly. It is incredibly complex and hard to follow.
thanks for the comment. will try to add more step by step text in future videos
I agree, as a newbie, not easy to follow the steps, I gave up.
@@Cedric00789 feel free to join the discord channel if u have any questions
How does it beat MCP?
it worked better than MCP when the video was recorded. However. MCP is always updating also, like it is now supporting a lot more LLM than it was launched. hence beats MCP might not be accurate as of now.
@@DevsKingdom Got it, thanks for the quick reply!
You should always clear the chat/start a new one otherwise you have context of old questions within the replies of unrelated ones.
thanks for the comment