|
π€ *Beep Beep - confident at ββββββ 99.9% certainty*π€ |
|
A collection of some of my favorite, interesting and practical repositories. Click to discover everything while exploring!
π£ A convolutional autoencoder that gets latent representations and evaluates them.
π©βπ» Use a temporal wrapper, featuring Gru-bi RRN, 3D convolutions, and sequence preprocessing.
π Includes L2 and cosine distance to effectively evaluate each language trained.
π³ Dockerized distributed cloud architecture.
π LangChain and Langraph to organize different agents communication.
π§ Uses RAG strategy with ChromaDB ingested with 2020 Tweets.
π Use an Gemini API to analice information.
π Include Maestro SDK for its deployment.
π‘ Has a nice web interface to showcase the app.
β Design in AWS using EC2 instances with custom RNN to generate a new scientific name.
π¦ Ollama integration to generate the physical description and Dall-e to visualize it, both via Ngrok.
π±βπ Immersive web museum to chat with any dinosaur created using the Gemini API.
πΆ Orchestatres serverless ETL pipeline using AWS Lambda to automate scraping and processing of news webs.
π² Integrates AWS Glue and Athena to catalog, transform, and store structured data Amazon S3.
π Generates machine-learning-ready datasets in CSV format for predictive analysis and modeling.
β Integrates simple cases of AI in Azure virtual machines.
β¨ Have different experiments, from object detection in images to extracting text from images.
β This repository is a copy of a set of official tutorials from Microsoft as a tutorial for the use of AI in cloud.
π Ingest data from Sakila in a AWS RDS utilizing ETL process as a introduction to Big Data.
π Uses AWS Glue Workflows to orchestrate the process executing jobs, then crawlers.
β Includes CI/CD workflows to accelerate deployment and ensure the robustness of the web application.
π Implements a RAG search over embeddings of text and images to recognize train cars.
β Evaluates the search with RAGAS, Faithfulness, Answer Similarity, and Context Precision.
π½ Uses a Gemini API as a multimodal tool to generate the final response with the results.
π Starts with basic concepts to advanced time series, data cleaning, interpretation, and predictive models.
π A collection of different notebooks that document years of learning Data Analisis and ML.
π Integrates various final projects with deep analysis of data in practical cases.
π₯ The collection goes from basic prediction models to deep data evaluation and complex clustering tasks.
β The final projects include experiments with supervised and unsupervised learning.
π Has methods of feature engineering, optimization, and model evaluation with tools such as the ROC curve.
