Update README.md

[update] README for project description

README.md

@@ -1,2 +1,42 @@
 # llama-supercazzola
-An educational project aimed at introducing high school students to the concept of neural networks through the use of "supercazzole" from Mario Monicelli's movies, "Amici Miei" (My Friends).
+
+This repository contains an educational project aimed at introducing high school students to the concept of neural networks through the use of gibberish lines from Mario Monicelli's movies, "Amici Miei" (My Friends).
+
+### Project Description:
+
+The "Supercazzole Neurali" project offers students a unique opportunity to explore the workings of neural networks and language models. Using a dataset containing lines of gibberish from the "Amici Miei" movies, the project allows students to understand how a neural network can be trained to generate text in response to specific inputs. The project leverages the llama-cpp framework for both inference and training tasks on llama models. It utilizes C++ exclusively, avoiding Python altogether.
+
+### Features:
+
+- Language model training and inference using llama-cpp framework.
+- Generation of gibberish in response to questions.
+- Expansion of knowledge on neural networks and natural language.
+- Convenient scripts for training and inference: `training.sh` and `inference.sh`.
+
+### Usage Instructions:
+
+1. Clone the repository to your local environment.
+2. Build the llama-cpp framework for training and inference tasks.
+3. Execute `training.sh` for training the model and `inference.sh` for an inference example.
+4. You can also utlize the model with the llama-cpp server.
+
+### Requirements:
+
+- C++ compiler and development environment
+- llama-cpp framework and dependencies
+
+### Credits:
+
+This project owes its feasibility and functionality to the exiting llama.cpp repository by [ggerganov](https://github.com/ggerganov/llama.cpp), which provides essential capabilities for neural network training and inference in C++.
+
+### Contributions:
+
+We welcome contributions and improvements from the community. Feel free to contribute to the project if you have any ideas.
+
+### Credits:
+
+This project was conceived and developed with love by [buzzobuono] for educational purposes.
+
+### License:
+
+This project is released under the [GPL-3.0] license. Refer to the LICENSE file for more information.