Photo for Author | Kanele
Obvious Introduction
This is the second title in my first project series. If you have never seen the first in Python, it is appropriate to check: The 5th sweetest Python projects for the perfect beginnings.
So, what's AI productive AI or Gen Ai? It is about making new content such as text, photos, code, sound, or even video using AI. Before the great language and model models, things were very different. But now, with the increase in basic models such as GPT, Llama, Nellava, everything is changed. You can create old tools and applicable apps without training models from the beginning.
I have chosen these 5 projects to cover a little bit all: Scripture, picture, voice, vision, and other backup concepts such as Tuning and RAG good rag. You will receive both API solutions and local sets, and by the end, you will have contacted all construction blocks used in modern general Gen Ai modern. So, let's get started.
Obvious 1. Recipe generator app (a generation generation)
Link: Create Recipe Generator with Reche: Code meets in the kitchen
We will start with something simple and fun that only uses the production of text and API key, no need for heavy setup. This app allows you to enter a few basic information such as ingredients, food type, food choices, cooking time, and difficulty. Then the perfect recipe issued the GPT. You will learn how to create a FRONTLEND form, send data to GPT, and give AI back to the user. Here is another advanced version of the same idea: create a AI recipes with GPT O1-Preview at 1 Hour. This is a high-quality engineering, GPT-4 recommendations, suggestions, ingredients, and Frontic Frontlend.
Obvious 2
Link: Create Python Ai Image Generator in 15 minutes (FREE & Location)
Yes, you can produce cool pictures using chatGPT tools, Dall · e, or Midjourney by typing. But what if you want to take action again and run everything without the API costs or the ends of the cloud? This project does exactly. In this video, you will learn how you can set up a stable generation on your computer. The Creator keeps simple: You put Python, Clone a Lightweight Ui Repo, Download the test model, and then run a local server. That's all. After that, you can enter the motivations on your browser and generate AI remains, everything except the Internet or API telephones.
Obvious 3. Vision Chatbot Vision + Vision + Text
Link: Create AI of AI LIYS app using Multimodal LLM LLM LLAVA and whispering
The project is not directly constructed as medical chatbot, but the case of the use fits well. You talk to her, listening, can look at the picture (such as X-ray or document), and responds to combining three ways: Word, vision and text. Designed for LLAVA (Multimodal model of language observation) and the Model of Backup (Opelai-to-to-T-TA-Text) of the Gristatio. The video moves in Colob, to include libraries, luxurious LLAVA to operate your GPU, and put it all and the GTTS for audio answers.
Obvious 4. Good planning of modern llms
Link: Fine Tune Gemma 3, QWEN3, Llan3, Llan3, Llama 4, Phila 4 and are very different and other ways
So far, we used off-the-sholf models with instant engineers. That works, but if you want to control more, good order is the next step. This video from Tzelis Research is one of the best out there. So, instead of lifting a project that just changes a good model, I was looking for you to focus on the actual configuration process. This video shows you how to know Gemma 3, QWEN3, Llan3, Llan3, Llan3, Llan3, Llan 4, Listening (Listening (Listening (Listening), how to prepare for Datasets, run fast of training, and training problems to reality.
Obvious 5. Build a local rag from the beginning
Link: Local Generation of Returns Light Measures (RAG) from the beginning (Step by Step Learning)
Everyone loves good Chatbot, but most falls outside when asked about things without their training information. This is where the rag is useful. You give your own llm of your VM Database of the correct documents, and it draws the context before answering. The video moves in creating a full area of the area using the Colab writing or your machine. You will upload documents (such as Texbook PDF), distinguishes the chunks, producing embeds, and linking it all in the local llm (eg LLAMA 2 for Ollama). It is the most clear rag lesson I have seen, and once you have done this, you will understand that Chatgpt plugins, AI search tools, and the internal company conversations are really effective.
Obvious Rolling up
Each of these projects teaching something important:
→ Image → Voice → Return → Route
When you just log in to Gen Ai and you want to actually build things, not just playing with demo, this is your Blueprint. Start from the most fun. And remember, it's ok to break things. It's the way you read.
Kanal Mehreen Kanwal is a machine learning device and a technical writer who has a great deal of data science and a combination of Ai and a drug. Authorized EBOOK “that added a product with chatGPT”. As a Google scene 2022 in the Apac, it is a sign of diversity and the beauty of education. He was recognized as a Teradata variation in a Tech scholar, Mitacs Globalk scholar research, and the Harvard of Code Scholar. Kanalal is a zealous attorney for a change, who removes Femcodes to equip women to women.
