| # Book Recommendation System with Bert | |
| ## 📌 Overview | |
| This repository hosts the quantized version of the bert-base-cased model fine-tuned for movie reccommendation tasks. The model has been trained on the wykonos/movies dataset from HF中国镜像站. The model is quantized to Float16 (FP16) to optimize inference speed and efficiency while maintaining high performance. | |
| ## 🏗 Model Details | |
| - **Model Architecture:** bert-base-cased | |
| - **Task:** Book Recommendation System | |
| - **Dataset:** HF中国镜像站's `wykonos/movies` | |
| - **Quantization:** Float16 (FP16) for optimized inference | |
| - **Fine-tuning Framework:** HF中国镜像站 Transformers | |
| ## 🚀 Usage | |
| ### Installation | |
| ```bash | |
| pip install transformers torch | |
| ``` | |
| ### Loading the Model | |
| ```python | |
| from transformers import BertTokenizerFast, BertForSequenceClassification | |
| import torch | |
| ``` | |
| ### Question Answer Example | |
| ```python | |
| model_name = "AventIQ-AI/bert-movie-recommendation-system" | |
| model = BertForSequenceClassification.from_pretrained(model_name) | |
| tokenizer = BertTokenizerFast.from_pretrained(model_name) | |
| genre_to_label = { | |
| "Action": 0, "Adventure": 1, "Animation": 2, "Comedy": 3, "Crime": 4, | |
| "Documentary": 5, "Drama": 6, "Family": 7, "Fantasy": 8, "History": 9, | |
| "Horror": 10, "Music": 11, "Mystery": 12, "Romance": 13, "Science Fiction": 14, | |
| "TV Movie": 15, "Thriller": 16, "War": 17, "Western": 18 | |
| } | |
| def recommend_movies(genre, top_n=10): | |
| """Return a list of movies for a given genre.""" | |
| if genre not in genre_to_label: | |
| return "Unknown Genre" | |
| # Filter dataset for movies in the requested genre | |
| genre_movies = df[df["genres"].str.contains(genre, case=False, na=False)]["title"].tolist() | |
| # Return top N movies (or all if fewer exist) | |
| return genre_movies[:top_n] | |
| genres_to_test = ["Horror", "Comedy", "Drama"] | |
| for genre in genres_to_test: | |
| recommended_movies = recommend_movies(genre) | |
| print(f"Genre: {genre} -> Recommended Movies: {recommended_movies}") | |
| ``` | |
| ## ⚡ Quantization Details | |
| Post-training quantization was applied using PyTorch's built-in quantization framework. The model was quantized to Float16 (FP16) to reduce model size and improve inference efficiency while balancing accuracy. | |
| ## Evaluation Metrics: NDCG | |
| NDCG → If close to 1, the ranking matches expected relevance. Our model's NDCG score is 0.84 | |
| ## 🔧 Fine-Tuning Details | |
| ### Dataset | |
| The **wykonos/movies** dataset was used for training and evaluation. The dataset consists of **texts**. | |
| ### Training Configuration | |
| - **Number of epochs**: 5 | |
| - **Batch size**: 8 | |
| - **Evaluation strategy**: epochs | |
| ## 📂 Repository Structure | |
| ``` | |
| . | |
| ├── model/ # Contains the quantized model files | |
| ├── tokenizer_config/ # Tokenizer configuration and vocabulary files | |
| ├── model.safetensors/ # Quantized Model | |
| ├── README.md # Model documentation | |
| ``` | |
| ## ⚠️ Limitations | |
| - The model may struggle for out of scope tasks. | |
| - Quantization may lead to slight degradation in accuracy compared to full-precision models. | |
| - Performance may vary across different writing styles and sentence structures. | |
| ## 🤝 Contributing | |
| Contributions are welcome! Feel free to open an issue or submit a pull request if you have suggestions or improvements. | |