5 ms·
Nice write up Sebastian, looking forward to the book. There are lots of details on the LLM and how it’s composed, would be great if you can expand on how Llama
by abusaidm 2y ago
Nice write up Sebastian, looking forward to the book. There are lots of details on the LLM and how it’s composed, would be great if you can expand on how Llama and OpenAI could be cleaning and structuring their training data given it seems this is where the battle is heading in the long run.
- rakahn 2y agoYes. Would love to read that.
- rahimnathwani 2y agohow Llama and OpenAI could be cleaning and structuring their training data If you're interested in this, there are several sections in the Llama paper you will likely enjoy: https://ai.meta.com/research/publications/the-llama-3-herd-of-models/ https://ai.meta.com/research/publications/the-llama-3-herd-o...
- kbrkbr 2y agoBut isn't it the beauty of llm's that they need comparably little preparation (unstructured text as input) and pick the features on their own so to say? edit: grammar
- aDyslecticCrow 2y agoYes, if you want an LLM that doesn't listen to instructions and just endlessly babbles about anything and everything. What turned GPT into chatGPT was a lot of structured training with human feedback.
- rahimnathwani 2y agoExactly. Section 4.3.7 briefly explains how they trained the model to better follow instructions ('steerability').