Llama 2 - The next generation of Meta's open source LLM

@chrismessina How is @chrismessina the hunter behind everything big? :) And would he hunt for us? We're building something cool too. XP Lover all your hunts man and it would be great engaging with you 🙏

@rajtilakjee how will you use this?

@chrismessina I am building a product which requires voice synthesis and text generation. So far I have been using Eleven Labs for voice synthesis, and GPT-3.5-Turbo for text generation. Eleven Labs was on a paid subscription since the free offer has a character limit of 2500, OpenAI was on the grant which when it will end will go into a GPT-4 subscription plan. This is mostly because I could not find an commercial open-source alternative till now. However, things are about to change. For voice synthesis I would be using Bark, it's the next best thing to Eleven Labs which is free and can be used commercially. And for text generation I am planning to use LLama2. Read some of the stats that LeCun share on Twitter and they looked promising. Got the approval for using the models last night (India time), now all I need is to refactor the code using HuggingFace API. Exciting times, huh? BTW, still cannot believe Chris Messina replied to my comment on PH.

@peppereddirt unclear how that applies here?

@cristinabigail77 yes: https://ai.meta.com/llama/#respo... "Llama-2-chat has undergone testing by external partners and internal teams to identify performance gaps and mitigate potentially problematic responses in chat use cases. We're committed to ongoing red-teaming to enhance safety and performance."

@seekrsi there's no waitlist. You just need to send them your email and you'll get access. I did.

@chris_patten_walker1 yup!

@chrismessina Not clear to me how to get access. Feels like I have do some installation in order to get things going. Correct?