Pinned Loading

1. KAN_Titanic KAN_Titanic Public

   Code to train and graphically represent de Titanic model with pykan (Python package) using KANs. Indeed, the whole idea of KANs promises to reduce the architecture complexity of models, compared to…

   Jupyter Notebook

2. Marcuswas-bert-drug-review-to-condition Marcuswas-bert-drug-review-to-condition Public

   Fine-tuning of a BERT-base-uncased model using HuggingFace Transformers library, on the dataset Drugs.com reviews (see below) with the purpose of text classification.

   Jupyter Notebook

3. Time-Series-PatchTST Time-Series-PatchTST Public

   I conducted an experiment following the ideas of Nie et al. (2003) https://arxiv.org/pdf/2211.14730.pdf However, I changed their setting, using NeuralForecast for predictions on the Exchange datase…

   Jupyter Notebook

4. RAG_talk_to_postgreSQL_database RAG_talk_to_postgreSQL_database Public

   This is a demonstration of how to set up an app through which you can use chatGPT to send text questions and fetch your data automatically from a postgreSQL database (Supabase)

   Python

5. SQL_BI_analyses SQL_BI_analyses Public

   What to do when you suspect a dip in engagement on your platform? What indicators can you look at? What kind of SQL queries can you perform to get clear evidence? What recommendations can you put f…

6. cross-encoders-and-re-ranking-in-RAG cross-encoders-and-re-ranking-in-RAG Public

   Deeplaerning.AI notebook from the course 'Advanced Retrieval for AI with Chroma" on cross-encoders and re-ranking

   Jupyter Notebook