- 177
- 1 158 401
Developers Digest
Registrace 17. 01. 2023
AI đ€ Development
Cursor: Revolutionizing Code Editing with AI
In this video, I revisit Cursor, a code editor originally covered on this channel about a year and a half ago. Cursor has evolved significantly since then, boasting a range of new features and capabilities. I provide a detailed demo showcasing its standout features, including its seamless transition from VS Code, its robust support for various code models, and its impressive AI-driven coding assistance tools. Key highlights include a mobile view demo, feature integration for improving UI elements, and efficient troubleshooting using AI. I also discuss Cursorâs pricing tiers and free trial options, making it accessible for both beginners and seasoned developers. Join me as I explore Cursor's potential to revolutionize the coding experience.
00:00 Introduction to Cursor
00:26 Cursor's Free and Pro Tiers
01:08 Setting Up a Next.js App with Cursor
01:52 Exploring Cursor's Features
04:03 Making Edits and Submitting Changes
07:01 Advanced Features and Commands
09:46 Final Thoughts and Recommendations
10:38 Conclusion and Future Plans
00:00 Introduction to Cursor
00:26 Cursor's Free and Pro Tiers
01:08 Setting Up a Next.js App with Cursor
01:52 Exploring Cursor's Features
04:03 Making Edits and Submitting Changes
07:01 Advanced Features and Commands
09:46 Final Thoughts and Recommendations
10:38 Conclusion and Future Plans
zhlĂ©dnutĂ: 5 717
Video
How to Fine-Tune Flux.1 in 3 Minutes: A Beginner's Guide
zhlĂ©dnutĂ 2KPĆed 14 hodinami
Fine-Tune Flux One: Create Hyper Realistic & Cartoon Images Easily! In this video, I demonstrate how to fine-tune Flux One, an advanced image generation model. Watch as I guide you through the process, from acquiring free credits on Fal.AI to uploading images, customizing settings, and generating your own images. Learn how to continually create consistent photos of various subjects, be it famou...
Prompt Caching: Anthropic Claude 3.5 Sonnet's Game-Changing Update!
zhlĂ©dnutĂ 2,2KPĆed 19 hodinami
In this video, we delve into Anthropic's latest release-prompt caching with Claude-right after Google's context caching announcement for Gemini 1.5. We explore the benefits of this technology, particularly for developers using Claude 3.5 Sonnet, and its potential to save both time and cost across various use cases. Discover how prompt caching can enhance applications like conversational agents,...
AI Text-to-Image-to-Video Guide: Quick Start Options in 4 Min
zhlĂ©dnutĂ 337PĆed 21 hodinou
In this video, I explore various AI tools available today for generating images and converting them into videos. I start by discussing the new open-source model, Flux One, accessible via Grok with a Twitter account or through Hugging Face. I then highlight other text-to-image models on Hugging Face, ChatGPT's free image generation, and the video outputs you can create using Kling AI, Luma Labs,...
Get Started with Upstage AI's Solar LLM in 10 Minutes
zhlĂ©dnutĂ 263PĆed dnem
Creating a Retrieval-Augmented Generation (RAG) Workflow with Upstage AI Upstage Console Link: go.upstage.ai/SOLAR_DEV_DIGEST Credit Redeem Coupon Code: SOLAR_DEV_DIGEST_2408 Start registering for the coupon: August 15, 2024, 12:00 AM (UTC) End of coupon registration: September 15, 2024, 12:00 AM (UTC) Credit expiration date: November 1, 2024, 12:00 AM (UTC) Credit amount: $30 Console: console....
Grok-2: xAI's New Frontier Model
zhlĂ©dnutĂ 1,3KPĆed dnem
In this video, I introduce the beta release of Grok 2 and Grok 2 Mini. I discuss the new models available on the X platform and their impressive performance, including their ranking on the Chatbot Arena and comparisons to other models like GPT-4 Turbo and Cloud 3.5 Sonnet. I also touch on Grok's focus on factuality, evaluation scores, and the exciting partnership with Black Forest Labs for imag...
OpenAI: New 100% Reliable Structured Outputs
zhlĂ©dnutĂ 7KPĆed 14 dny
đ OpenAI's Game-Changing API Update: 100% Reliable JSON Outputs Explained! In this video, we delve into OpenAI's latest API update introducing structured outputs. We explore how it differs from JSON mode by ensuring outputs conform to provided schemas, and discuss its implications for developers. Key topics include function calling, response format parameters, and safety measures. We also cover...
Web Scraping to SMS Alerts with Groq/Ollama Llama 3.1 + Twilio
zhlĂ©dnutĂ 659PĆed 14 dny
Create a Powerful AI-Driven News Alert System with OpenAI and Twilio! In this video, learn how to set up an AI notification system by writing a simple script that scrapes Hacker News using Puppeteer, filters relevant news about AI and software development, and sends notifications via Twilio. The tutorial walks you through the entire process, including the setup and configuration of OpenAI, Grok...
Deploy ANY Open-Source LLM with Ollama on an AWS EC2 + GPU in 10 Min (Llama-3.1, Gemma-2 etc.)
zhlĂ©dnutĂ 1,6KPĆed 21 dnem
In this video, I demonstrate how to set up and deploy a Llama 3.1 Phi Mistral Gemma 2 model using Olama on an AWS EC2 instance with GPU. Starting from scratch, I guide you through the entire process on AWS, including instance setup, selecting the appropriate AMI, configuring the instance, and setting up the environment with CUDA drivers. We also cover installing Go, cloning a simple Go server, ...
Did Gemini 1.5 Pro Just Beat GPT-4o?
zhlĂ©dnutĂ 1,4KPĆed 21 dnem
In the latest video, we discuss Googleâs recent advancements in the AI field, particularly focusing on the release of Gemini 1.5 Pro and Gemma 2 models. We examine their impressive performance on the LMSYS Chatbot Arena leaderboard, noting how they surpass competitors like GPT-3.5 and Lama 2 70B based on human feedback. An overview of the Chatbot Arena evaluation process and its emphasis on use...
Ollama Function Calling: LangChain & Llama 3.1 đŠ
zhlĂ©dnutĂ 2KPĆed 21 dnem
In this video, I introduce the tool support feature within Ollama that enables local function calling on your machine. We'll explore the easy setup, deployment options (local and cloud), and how to leverage different models, notably Llama 3.1 and Groq (function calling model). I'll demonstrate practical examples, including weather queries for San Francisco, opening applications like the calcula...
Go + HTMX + OpenAI: Create a Lightweight AI Chat Application
zhlĂ©dnutĂ 793PĆed 21 dnem
Building an Efficient LLM Chat Application with Go and HTMX In this video, I will guide you through building a large language model (LLM) chat application using Go and HTMX. Inspired by fellow CZcamsr Web Dev Cody's frustrations with TypeScript and Next.js, I'm exploring how these alternatives can offer more resource-efficient solutions. I'll compare memory usage between a Next.js and Go-based ...
SearchGPT: OpenAI's New AI Search Prototype
zhlĂ©dnutĂ 19KPĆed mÄsĂcem
In this video, I dive into SearchGPT, OpenAI's new AI search prototype designed to deliver fast and accurate answers with clear and relevant sources. I discuss its user interface, features like auto-complete, and how it references sources in-line and through a panel. Additionally, I touch on the partnerships OpenAI is forming with publishers to ensure fair representation of their content and ho...
Fine-Tuning GPT-4o with Synthetic Data: A Step-by-Step Guide
zhlĂ©dnutĂ 7KPĆed mÄsĂcem
In this video, I will show you how to fine-tune GPT-4.0 mini to achieve better results by generating synthetic data using Claude 3.5 Sonnet. We'll cover how to dynamically create SVGs, normalize the data, and set up a fine-tuning session using the OpenAI platform. Additionally, we'll discuss setting up a simple script to generate these SVGs, perform human intervention to select the best ones, a...
Mistral Large 2 in 4 Minutes
zhlĂ©dnutĂ 3,3KPĆed mÄsĂcem
In this video, I review the latest model from Mistral, the Mistral Large 2. I'll discuss its impressive 128,000 context window, its training on dozens of languages including more than 80 coding languages, and its performance benchmarks. Comparing it with models like LLAMA 3.1 and GPT 4o, Mistral Large 2 stands out in various coding tasks and instruction following. I also cover its licensing opt...
Llama 3.1 405B REWRITES Open Source Frontier
zhlĂ©dnutĂ 697PĆed mÄsĂcem
Llama 3.1 405B REWRITES Open Source Frontier
GPT-4o Mini: OpenAI's Multimodal Marvel UNVEILED
zhlĂ©dnutĂ 1,7KPĆed mÄsĂcem
GPT-4o Mini: OpenAI's Multimodal Marvel UNVEILED
Micro Agent: Create Functions with Natural Language in Your Terminal
zhlĂ©dnutĂ 781PĆed mÄsĂcem
Micro Agent: Create Functions with Natural Language in Your Terminal
Moshi: The Open-Source Answer to GPT-4's Voice Model?
zhlĂ©dnutĂ 2KPĆed mÄsĂcem
Moshi: The Open-Source Answer to GPT-4's Voice Model?
I Spent $15 on Runway Gen-3 Alpha AI Videos: Worth It?
zhlĂ©dnutĂ 11KPĆed mÄsĂcem
I Spent $15 on Runway Gen-3 Alpha AI Videos: Worth It?
Hume AI: Empathic Voice Interface - Next.JS Template
zhlĂ©dnutĂ 1,5KPĆed mÄsĂcem
Hume AI: Empathic Voice Interface - Next.JS Template
Gemma 2 Aims for the Crown! Google's Latest Open-Weights 9B & 27B Models
zhlĂ©dnutĂ 1,3KPĆed mÄsĂcem
Gemma 2 Aims for the Crown! Google's Latest Open-Weights 9B & 27B Models
Build No-Code AI Applications with Dify Drag-and-Drop Workflows
zhlĂ©dnutĂ 2,9KPĆed mÄsĂcem
Build No-Code AI Applications with Dify Drag-and-Drop Workflows
Building Anthropic Claude 3.5 Sonnet Artifacts: Step-by-Step Tutorial
zhlĂ©dnutĂ 2,7KPĆed 2 mÄsĂci
Building Anthropic Claude 3.5 Sonnet Artifacts: Step-by-Step Tutorial
Claude 3.5 Sonnet đ: Anthropic's Answer to GPT-4o
zhlĂ©dnutĂ 2,2KPĆed 2 mÄsĂci
Claude 3.5 Sonnet đ: Anthropic's Answer to GPT-4o
How to Self Host Coolify on AWS EC2: Step-by-Step Guide
zhlĂ©dnutĂ 2,3KPĆed 2 mÄsĂci
How to Self Host Coolify on AWS EC2: Step-by-Step Guide
Open WebUI: Self-Hosted Offline LLM UI for Ollama + Groq and More
zhlĂ©dnutĂ 7KPĆed 2 mÄsĂci
Open WebUI: Self-Hosted Offline LLM UI for Ollama Groq and More
Amazon Bedrock: Integrating Multiple LLMs
zhlĂ©dnutĂ 462PĆed 2 mÄsĂci
Amazon Bedrock: Integrating Multiple LLMs
Codestral: Mistral AI's FIRST Coding Model
zhlĂ©dnutĂ 4,7KPĆed 2 mÄsĂci
Codestral: Mistral AI's FIRST Coding Model
NOT WORTH IT đ€ Real SCAM
Fastest and better than copilot
The composer features is genius. It update all files at once instead to ask Claude to do it one by one
How do I get that composer view?
cmd + i / cmd + shift + i (for full screen) Iâd imagine it would be a similar cmd on windows
Thank you for your vedio! Can i train flux lora model twice? I trained the flux model with 10 images with a certain keyword and now wanted to retrain it with a different set of images and with a different trigger word. Any idea if that is possible?
I believe the cache can only be updated 4 times per cache key
best channel!
đ â€ïž
not clear for me what Free tier includes; the 2000 completions + 50 slow premium requests + 200 cursor-small uses are for a month (refilled in every month) or one time (if any of list depletes then free options is over)
youtubers really be giving out the sauce, thats crazy
I tried this but ultimately went back to VScode. The DX is ways better with a VScode plugin like cody, especially if you want to run your own LLM. For some reason they made that really hard to do in cursor.
wheres the fkin github link dummy!
Which AI tooling can one consider, if i want to migrate a rather large MERN stack codebase to NextJS?
This is a great high-level overview. Thank you so much for putting this video together.đ
Thanks for watching and all the great questions!
The other thought I have is when you have really large files itâs helpful to reactor them into smaller files that are more manageable and that refactoring is some thing that seems to be easier to do with copy paste rather than working with specific files that you already have in focus. I wonder how cursor helps Refactoring of a very large file into multiple smaller files without leaving their user interface?
I have tested this in the composer view - it does have the ability to create new files + directories for you which could help for refactors like this
I love their application I just wish I could use my own API keys AND not buy one of their plans. Iâm assuming their code does the interleaving for you so that you donât have to get the entire code file from Sonnet 3.5 their application does the interleaving of code which means that the amount of tokens that their application uses is probably very low.
docs.cursor.com/miscellaneous/api-keys#custom-api-keys You can add your custom keys, here are some of the options if you'd like!
You may want to check out Zed AI.
Once you accept the changes, do you lose your ability to undo the work? I guess you could use a git push to save your work before you start an iteration, but a lot of times I like to do and undo if Sonnet 3.5 breaks the code using visual studio code to quickly revert back to a known good state Even before doing a git push.
You can undo - but it never hurts to commit your work... I would be mindful of this for beta features like composer just to be safe though
first comment!!
Outstanding! You just saved my life; I owe you a coffee. đ
Nice, but I would like to do it in my own pc
hi there, nice tutorial, one question, Why are you repeating the same system role? cant you just have 1 system role and multiple examples of user and assistant?
Claude is getting stupid (its being quantized) To bad.
what does that mean?
Claude is gone dumb and forgetful.
how do we run this project :)
I want to get pictures of realistic shoes, can I finetune with different images of shoes model even though they look different?
This is absolutely incredible!!
Would be interesting to see if it's possible to have a 'shared cache' ie. for a common system prompts across a multi-user LLM implementation where you have a big system prompt that you want to be common across all sessions. Surely something like this is in the work or already exists. Can you "fork" a chat with a cached prompt? Otherwise it might not always be worth caching the prompt with the extra pricing if the user only asks one or two questions.
this is the easiest offline chatbot ive used zero coding, zero command propmpts to mess with this is beginner friendly. AND YOU CAN TALK TO ENTIRE BOOKS AND HAVE IT SUMMARIZE THEM FOR YOU ITS ALL FOR FREEEEEEE
Great video!
I've been struggling with getting GPT to give me solid JSON, this is a WELCOME update!!! thanks for the great video description too.
Thanks for watching!
hey, love your videos. But this running text in the middle is annoying. Please keep your usual style of the video. Or at least, don't put something that jumps so much.
Super cool!
How can I run the javascript code in a browser console?
this is server-sided, any logs would show up in the terminal if you wanna log things in your browser console. one way of doing it is adding a endpoint that sends the data to a frontend then the frontend logs that data, but usually it's not a good idea since you use backends for security
thanks for your videos, my man, you are a real hero.
â€ïžđ
awesome video, please don't stop!
Thank you! đ
I am curious if Runway Gen 3 can create a character. I was looking to create a character with a realistic face and use it in many CZcams videos. Not sure if Runway can do this yet.
maybe a dumb question. how do you turn the stream data you received into readable sentences
You could accumulate tokens and split by the end of sentences . ! ? Etc and then send resp after grouping function like that
I tried Gen3 too...personally, not worth it: Gen3: 15$ month ,=12 generations. Luma: 10$ =70 !!! Gen3 is probably.... Âżmicroscopically better? NO contest, IMO
You should check out Uizard
Can you use open web ui?
It's too expensive and the quality is hit or miss. When it hits, it's magical, but the misses are always so disappointing (like, it'll just have ignored parts of what i told it). If Runway can make it cheaper, then I'm fine to keep plugging away hoping to get a happy accident. Until then, its overpriced.
i guess, this is a javascript example
No work interpretor Local
I was searching to build a hume like evi ai for my project, but everyone was just talking about it was great so and so, finally found a good video, keep up this work and please put a video on how to create a evi ai on localhosy
So epic!đ Any idea when the code repo for this project will be out ? Or if anyone knows any other currently available ?
man, tysm!
Thanks for watching!
Hi, can we integrate OpenAI TTS / deepgram here instead of webkit?
I found that fine tuning a model specifically to handle output produces the most reliable results... this might change things. Thanks for the update!
I read this blog post but I should have just watched this. So much insightful context!
Ty! đ
Give whoever pushed forward on that project a raise. Really high impact.
I agree đ
the voice in this video is so irritating. why did you use this kind of voice. sorry, but really, it sounds distorted, irritating and annoying. cant you reacord in a neutral and clear voice ??
I stumbled upon this update just now when in the playground UI. Why is this announcement halfway down the page? This is so important! đ Since September of last year, I've been abusing function calling mode to do this kind of thing for my own framework. It's so messy! Having Pydantic integration and pre-parsed return values will make my code so much cleaner! ...guess I'm spending time tomorrow after work refactoring half my framework's main class methods đ
đ