https://www.theverge.com/2024/4/18/24133808/meta-ai-assistant-llama-3-chatgpt-openai-rival What do you all think about Llama 3 and how it compares with Chat GPT -- is it actually a possible candidate for Meta to win the AI race? Anyone use both Chat GPT and Llama 3 extensively to comment on how they compare?
Well llama 2 is relatively weaker than GPT4. Llama3 is very new, so no one knows the answer yet.
Only competent answer
And OpenAI is ahead of the game so no one even knows what's up their sleeve. Basically they're shaping the road and everyone else is just trying to run on it
Llama 3 just came out so noone has a good opinion on it yet
I’ve heard the new LLAMA was trained using data from Blind so it a little bit of a brat!
Probably lies about TC then
It’s 82 MMLU, still slightly worse than 84-86 from GPT4.
The 400B model has 84.8 MMLU but it's not finished training yet. See the end of https://ai.meta.com/blog/meta-llama-3/
Anthropic's Claude has been giving ChatGPT a run for its money! Personally, I have been using the free Claude over the free ChatGPT.
This. IMHO. Claude is much more accurate than Chat GPT for complex ambiguous prompts
I tried both an think ChatGPT 4 is still better
I’m a paid ChatGPT boi but keep finding myself using Claude instead
🥱
Anthropics Claude is better than all of them
If anything Anthropic is closer in the race
Go home Meta. You're drunk.