Meta has silently rolled out NotebookLlama, the open-source version of Google's NotebookLM. This Meta project uses Meta's Llama models for processing.

This is not the first time an attempt has been made to replicate Google's NotebookLM. However, some attempts were more successful than others. Meta is taking strides in the open-source AI world. So, by releasing an open version of NotebookLM, the tech giant is allowing public access to coding and documentation powered by AI.

NotebookLlama is powered by Meta's Llama language models which were made for code and interactive document generation. This open-source model makes a transcript right from the file, for example, a blog post. After which it then adds interruptions and dramatization ahead of sending the transcript to the open source models. However, the results from Meta's open source were not good as compared to NotebookLm. Meta has acknowledged this and the researchers working on the project have said that the quality will see improvement with stronger models.

NotebookLlama: More on this new release by Meta

Meta's researchers have stated that the text-to-speech model serves as a limitation for the results sounding natural. Furthermore, a different approach for the writing process of the podcast might be to have two agents debating on the topic of interest and write the podcast outline accordingly. However, due to certain setbacks, he researchers currently use only one model to write the outline of the podcast.

Meta's model enables developers to make more personalised models tailored to the specific needs of their projects. The tech giant has contributed a foundational model to apply NotebookLlama in different environments. The model also has support for multi-turn conversations which allow for much more in-depth interactions between the AI and the user.

Suggested:

AI Studio: Meta’s New AI Tool Has Been Launched in The US

Llama 3.1: Meta’s Biggest Open Model Has Been Released

Image Credits: Screenshot Taken From NotebookLlama Github