A very capable chat model built on top of the new Mistral MoE model, trained on the SlimOrca dataset for 1 epoch, using QLoRA.
A very capable chat model built on top of the new Mistral MoE model, trained on the SlimOrca dataset for 1 epoch, using QLoRA.
…
I’m going to go ahead and say this had a negative effect on the crew… There’s nothing quite like dangling a treat in front of someone, then snatching it away. Those poor astronauts.