Movie Recommender System in Python with LLMs
HTML-код
- Опубликовано: 28 сен 2024
- Today we build a movie recommender system using LLMs and vector stores in Python.
Ollama: www.ollama.com/
Dataset: www.kaggle.com...
◾◾◾◾◾◾◾◾◾◾◾◾◾◾◾◾◾
📚 Programming Books & Merch 📚
🐍 The Python Bible Book: www.neuralnine...
💻 The Algorithm Bible Book: www.neuralnine...
👕 Programming Merch: www.neuralnine...
💼 Services 💼
💻 Freelancing & Tutoring: www.neuralnine...
🌐 Social Media & Contact 🌐
📱 Website: www.neuralnine...
📷 Instagram: / neuralnine
🐦 Twitter: / neuralnine
🤵 LinkedIn: / neuralnine
📁 GitHub: github.com/Neu...
🎙 Discord: / discord
Although I have done similar projects, including with FAISS, I certainly did learn a lot from this exercise. As you suggested, some trial-and-error runs may be needed to determine if the best results should be based upon 'description', 'genre', etc. Of course, the major factor is the LLM used for embeddings. I used 'llama3.1 8b' which is usually very good, but was somewhat disappointing in this case. I will have to try combinations of data columns to embed, embedding models and other variations. BTW: I realized that much processing and "wall-clock" time can be saved by only processing the first 100 movies to validate the code and verify the results. Later, I can process all of the 3,000 or so movie records while I go to breakfast or get some sleep ! Thank You !!!
bro, you're so unbelievably unrated, please keep uploading videos even if all your vids views are not even 10k. Im a new subscriber
Another timely and very useful project. Thank You!
It would really be helpful to know what it was using to come up with the results.
Would you please provide the code for this project ? You didn`t upload on github also.
Nice topic
Do i need to wait for all instances to processed? I reach to almost 300 instances and its still going on....Help me
Did you use the base model or the instruct fine tuned?
How do I open Netflix titles cvs in pycharm?
Would you please provide the code for this project, please?
❤
Thx_.
Hi
legend
Can we do the same without using an LLM?
With code, you can do anything you like.
And now I really want to see this movie: "A group of adventurers discover a mysterious programming snake in the jungle and find something extremely shocking." 😁
my question, for what reason did you use llama2 for embedding and not, for example, the nomic-embed-text model?