Join the conversation

Join the community of Machine Learners and AI enthusiasts.

Sign Up
clemΒ 
posted an update Apr 21
Post
2906
Already almost 1,000 llama3 model variations have been shared publicly on HF (many more in private use at companies): https://huggingface.co/models?p=5&sort=trending&search=llama3.

Everyone should fine-tune their own models for their use-cases, languages, industry, infra constraints,...

10,000 llama3 variants by the end of next week?

The exploration space with the base model is huge, especially given that the 8B one can run on almost any hardware. My hunch is it will continue to prove that aside from compute budgets, data is everything. And how you creatively apply said data.

But our community is facing a dataset shortage.

Β·

When searching "llama-3" there's 3,588 πŸ™€