Mahesh Sathiamoorthy(@madiator):I am reading the exact opposite way. This effort is very cool, and isn't it kind of amazing to collect your own data, run a custom midtrain, and then running your own RL to end up at the same performance as Astra max for 1/2 the inference price? Imagine what can then be next.

Mahesh Sathiamoorthy

@madiator

CEO @bespokelabsai. RL/Envs/Posttraining/Agents. Created Generative Retrieval at @GoogleDeepMind.

加入 February 2008

1.5K 正在关注 17.3K 粉丝

Mahesh Sathiamoorthy@madiator

2026.09.16 14:09

I am reading the exact opposite way. This effort is very cool, and isn't it kind of amazing to collect your own data, run a custom midtrain, and then running your own RL to end up at the same performance as Astra max for 1/2 the inference price? Imagine what can then be next.

显示更多