1/ Six models, from 0.9B to 375B total parameters, released as one connected fleet.
@IFM_AI’s K2 Horizon is ambitious because it tries to make the training process inspectable, from pretraining to reasoning and agentic post-training.
A closer look:
1/ 六个模型,总参数量从 0.9B 到 375B,作为一个互联舰队发布。 @IFM_AI 的 K2 Horizon 之所以雄心勃勃,是因为它试图让训练过程可检视,从预训练到推理及智能体后训练。 细看:
1/ 六个模型,总参数量从 0.9B 到 375B,作为一个互联舰队发布。 @IFM_AI 的 K2 Horizon 之所以雄心勃勃,是因为它试图让训练过程可检视,从预训练到推理及智能体后训练。 细看:
1/ Six models, from 0.9B to 375B total parameters, released as one connected fleet.
@IFM_AI’s K2 Horizon is ambitious because it tries to make the training process inspectable, from pretraining to reasoning and agentic post-training.
A closer look:
来源:Chubby♨️· x.com