Miguel Ángel Medina Ramírez's repositories
Personal-Data-Protection-Service
Web service oriented to the location and protection of personal data in documents
ReID-in-Sport-competitions
Research about ReID in sport competitions using deep learning methods
Improving-Dialogue-Management
A detailed analysis of task-oriented dialogue systems, emphasizing the importance of high-quality datasets in dialogue management. We crafted a synthetic dialogue generator to highlight the impact of dataset errors on model outcomes.
MatLibraryPy
matrix library made with C ++ for Python
Analysis-of-tweets
Sentiment analysis of airline tweets
fake-face-detector
Experiments for the detection of false faces generated by GANs and deepfakes
alignment-handbook
Robust recipes to align language models with human and AI preferences
Car-detection-for-Autonomous-Driving
Car detection using yolo
Comment-Sentiment-Analysis
Sentiment analysis on an amazon food dataset.
Face-Recognition
A simple implementation for face recognition using convolutional networks
FinetuningTutorial
Finetuning tutorial for GPT-3
Heart-Failure-Prediction
Artificial intelligence for heart failure detection
lit-gpt
Hackable implementation of state-of-the-art open-source LLMs based on nanoGPT. Supports flash attention, 4-bit and 8-bit quantization, LoRA and LLaMA-Adapter fine-tuning, pre-training. Apache 2.0-licensed.
memformer
Implementation of Memformer, a Memory-augmented Transformer, in Pytorch
miguel-kjh.github.io
My personal website
NSGAII-Algoritm
Example of use of this multimodal genetic algorithm
PEDP
“Think Before You Speak”: Improving Multi-action Dialog Policy by Planning Single-Action Dialogs (IJCAI-22 long oral presentation)
PPO-deep-reinforcement-learning
An application of the PPO algorithm for specific environments
storm
An LLM-powered knowledge curation system that researches a topic and generates a full-length report with citations.
TODS-datasets-processor
Tool for downloading, managing and cleaning datasets for Task-Oriented Dialogue Systems.