1MLET_FASE4_TECH_CHALLENGE

Projeto que desenvolve um modelo LSTM para prever preços de ações. Inclui coleta de dados, criação e treinamento do modelo, deploy de uma API para previsões e monitoramento em produção. Entregáveis: código-fonte, documentação e scripts para deploy.

Deep Learning Project

This project utilizes a deep learning model implemented with PyTorch to perform data analysis and prediction. The project is organized modularly, with separate components for data collection, preprocessing, model definition, evaluation, and auxiliary functions.

Project Structure

General Structure

• main.py: Main script that executes the entire project pipeline.
• mlflow_setup.py: Configures MLflow for experiment tracking and logging.

Directories

data/

Scripts for data collection and preprocessing.

• fetch_data.py: Collects the necessary data for the project.
• preprocess_data.py: Processes the collected data, preparing it for the model.

models/

Scripts for defining, training, and saving the model.

• model_pytorch.py: Defines the deep learning model architecture using PyTorch.
• save_model.py: Contains the function to save the trained model.
• saved/: Stores the saved trained model (trained_model.pth).

predict/

Scripts for making predictions and evaluating model performance.

• evaluate_model.py: Evaluates the model's performance with relevant metrics.
• predict_pytorch.py: Uses the trained model to make predictions.

utils/

Auxiliary functions for various tasks across the project.

• device_utils.py: Manages device configurations, such as setting up a GPU.
• plot_utils.py: Contains functions for data visualization and plotting.
• sequence_utils.py: Handles data sequence manipulation and creation.
• tensor_utils.py: Utilities for tensor operations and manipulation.

---

How to Run

1. Install all necessary dependencies:

   pip install -r requirements.txt

2. Configure parameters (param) in src/ml/main.py as needed.
3. Execute the main pipeline using:

   python src/ml/main.py