No one need super expensive Macs to try 405B, Used server motherboard with 12 Ram slots costs $150. Llama models from Meta are always very boring to me, i've tried locally q8 quality 405b, same boring super censored model as usual, for request to write game code on it's choice it offered me textual no graphics rpg, what a disappointment.
So efficient and useful!
Very informative
o1 models are great in benchmarking I wonder why I still get better results in real life using Sonnet 3.5:)
To stay sane, I've decided to watch this update once and cut the rest of the AI Noise updates.
Go full screen on the slides!!
No one need super expensive Macs to try 405B, Used server motherboard with 12 Ram slots costs $150. Llama models from Meta are always very boring to me, i've tried locally q8 quality 405b, same boring super censored model as usual, for request to write game code on it's choice it offered me textual no graphics rpg, what a disappointment.
Orion is coming out in December,