This repository is a curated collection of Jupyter Notebooks focused on practical AI, NLP, and prompt engineering tasks I've managed to collect from excercises from various courses and online sources.. Each notebook demonstrates a specific concept, workflow, or experiment, organized by topic and toolset.
A collection of notebooks demonstrating the use of Hugging Face Transformers and related libraries for both NLP and CV tasks.
-
automatic-speech-recognition.ipynb
Example(s) of automatic speech-to-text using Hugging Face models. -
deployment.ipynb
Guides or experiments for deploying ML models, likely using Hugging Face tools. -
image-captioning.ipynb
Generating captions for images using pre-trained models. -
image-retrieval.ipynb
Retrieving images based on text queries or vice versa. -
nlp.ipynb
General NLP tasks using Hugging Face (tokenization, classification, etc.). -
object-detection.ipynb
Object detection in images using Hugging Face or compatible CV models. -
segmentation.ipynb
Image segmentation tasks. -
sentence-embeddings.ipynb
Generating and using sentence embeddings. -
text-to-speech.ipynb
Converting text input to speech output. -
translation-summarization.ipynb
Translation and summarization examples. -
visual-QnA.ipynb
Visual Question Answering: answering questions about images. -
zero-shot-audio-classification.ipynb
Audio classification without fine-tuning for specific classes. -
zero-shot-image-classification.ipynb
Image classification without fine-tuning.
Notebooks for constructing, expanding, and querying knowledge graphs, especially for retrieval-augmented generation (RAG) use cases.
-
0-prep-text-embeddings-for-RAG.ipynb
Prepares text embeddings for use in RAG pipelines. -
1-construct-kg-from-text-files.ipynb
Builds knowledge graphs from unstructured text. -
2-add_relationships_to_kg.ipynb
Augments graphs with relationships. -
3-extra-context-data-expand-of-kg.ipynb
Adds extra context data to the KG. -
4-chat-with-kg.ipynb
Interfaces for chatting or querying the KG. -
query-with-cypher.ipynb
Querying knowledge graphs using Cypher (Neo4j).
3. langchain/
Demonstrations and how-tos for using LangChain, a framework for developing LLM-powered applications.
-
agents.ipynb
Working with autonomous agents. -
chains.ipynb
Chaining multiple LLM operations. -
evaluation.ipynb
Evaluation strategies for LLM outputs. -
memory.ipynb
Handling conversational memory with LangChain. -
model-prompt-parser.ipynb
Parsing and constructing model prompts. -
qna-over-documents.ipynb
Question Answering on unstructured documents.
4. openai-api/
Hands-on tutorials and experiments using the OpenAI API.
-
basics-setup.ipynb
Getting started with the OpenAI API. -
chain-of-thought.ipynb
Demonstrations of chain-of-thought prompting. -
chaining-prompts.ipynb
How to chain prompts for complex workflows. -
check-outputs.ipynb
Inspecting and validating LLM outputs. -
classification.ipynb
Text classification using OpenAI models. -
end-to-end-example.ipynb
Complete workflow(s) from prompt to deployment. -
evaluation-1.ipynb, evaluation-2.ipynb
Evaluation of outputs, possibly using metrics or human-in-the-loop. -
moderation.ipynb
Content moderation with OpenAI models.
Best practices, experiments, and guides on prompt engineering for LLMs.
-
chatbot.ipynb
Building chatbots with prompt engineering. -
expanding.ipynb
Techniques for expanding prompts. -
guidelines.ipynb
General guidelines for writing effective prompts. -
inferring.ipynb
Inference tricks and techniques. -
iterative-prompt-development.ipynb
Iterative prompt development methodologies. -
summarizing.ipynb
Prompt engineering for summarization tasks. -
transforming.ipynb
Prompt engineering for data transformation tasks.
- Clone this repo:
git clone https://github.com/asleekgeek/notebook-collection.git - Open notebooks in your preferred Jupyter environment (JupyterLab, VSCode, Colab, etc.).
- Review the notebook for required dependencies (often listed at the top of each notebook).
- Run and experiment!
Contributions are welcome! Feel free to submit pull requests for new notebooks, improvements, or bug fixes.
This project is licensed under the MIT License.
Maintained by asleekgeek.
Notebook contributions and inspirations from the open-source AI/ML community.
Notebooks are for educational and experimental purposes. Some may require API keys or paid resources.