This repository provides a collection of Google Colab notebooks covering vision-language models, large language model (LLM) techniques such as RAG, quantization, fine-tuning, and computer vision applications. It is designed for developers, researchers, and AI enthusiasts to quickly explore and run state-of-the-art models directly in-browser — with no local setup required.
| Notebook | Description | Launch |
|---|---|---|
| Text-to-Video with WAN | Generate realistic AI video using the Wan2.2-T2V-A14B text-to-video model. |  |
| Quantize SeaLLM 7B Chat with ExLlamaV2 | Quantize and run the SeaLLM 7B chat LLM efficiently with ExLlamaV2. Includes a quantized model. | |
| Run LLaMA 3.2 Vision Model with Ollama | Deploy and run the LLaMA 3.2 Vision-Language Model locally with Ollama on Colab. | |
| Finetune Falcon 7B for Mental Health | Fine-tune the Falcon 7B LLM with QLoRA on a mental health chatbot dataset. Includes a finetuned model. | |
| Build Knowledge Graph with LlamaIndex | Construct a knowledge graph from documents using LlamaIndex for retrieval-augmented generation (RAG). | |
| Train Lipstick Detector | Train a cosmetics-focused computer vision model for lipstick detection using Mediapipe Model Maker. | |
- Click Open in Colab to launch the notebook.
- Sign in with your Google account.
- Run each cell in order to reproduce results.
- (Optional) Save a copy in your Drive for modifications.
Contributions are welcome!
- Open an Issue for bugs or suggestions.
- Submit a Pull Request for new notebooks or improvements.
Dieu Sang Ly — lydieusang@gmail.com
GitHub: @lydieusang