Ai2’s model, called Tulu 3 405B, also beats OpenAI’s GPT-4o on certain AI benchmarks, according to Ai2’s internal testing. Moreover, unlike GPT-4o (and even DeepSeek V3), Tulu 3 405B is open source, which means all of the components necessary to replicate it from scratch are freely available and permissively licensed.
You must log in or register to comment.
I thought DeepSeek was open source?
Nope, just the model weights.
Btw, neither is this model. And it even has a worse license, and doesn’t seem to compare at all, since this is a fine tune of Llama3.1 on maths questions?! and deepseek v3 is from the grounds up, has a MoE… And OpenAIs models are are different story altogether. Plus these benchmarks aren’t even far off… Or say anything (like most AI banchmarks).
