I think people are just happy that OpenAI is getting shit on, even if the reality isn’t really what is being portrayed. For example I’ve been trying to use r1-32b and it’s really no where near as good as Claude sonnet 3.5 has been.
I stopped using openai so I can’t comment on the performance comparison there, but clearly the benchmarks are all just made up bs.
I think another factor is that apparently the training of this LLM was done significantly cheaper than other mainstream models. Or that’s what I came across from other forum discussions. I don’t really care too much to bother myself and dig deeper
I think people are just happy that OpenAI is getting shit on, even if the reality isn’t really what is being portrayed. For example I’ve been trying to use r1-32b and it’s really no where near as good as Claude sonnet 3.5 has been.
I stopped using openai so I can’t comment on the performance comparison there, but clearly the benchmarks are all just made up bs.
I think another factor is that apparently the training of this LLM was done significantly cheaper than other mainstream models. Or that’s what I came across from other forum discussions. I don’t really care too much to bother myself and dig deeper