The curious case of LLama: How a leaked model sparked an open source AI revolution

In this video, we discuss Llama an LLM developed by Meta that claims performance close to monolithic GPT-3. Llama has given birth to instruction finetuned open source models that are challenging ChatGPT. Meta's official announcement: https://ai.facebook.com/blog/large-language-model-llama-meta-ai/?ref=the-batch-deeplearning-ai Llama paper: https://arxiv.org/pdf/2302.13971.pdf Instruct GPT paper: https://arxiv.org/pdf/2203.02155.pdf Sparks of AGI (GPT-4 demo) paper: https://arxiv.org/pdf/2303.12712.pdf Chinchilla paper: https://arxiv.org/pdf/2203.15556.pdf RoPE paper: https://arxiv.org/pdf/210…