Baidu: ernie-4.5-21b-a3b-pt:free
A sophisticated text-based Mixture-of-Experts (MoE) model featuring 21B total parameters with 3B activated per token, delivering exceptional multimodal understanding and generation through heterogeneous MoE structures and modality-isolated routing. Supporting an extensive 131K token context length, the model achieves efficient inference via multi-expert parallel collaboration and quantization, while advanced post-training techniques including SFT, DPO, and UPO ensure optimized performance across diverse applications with specialized routing and balancing losses for superior task handling.
Todos os provedores deste modelo estão ocupados agora
Cada provedor upstream atingiu seu limite de taxa. O modelo volta automaticamente quando os limites caem, geralmente em questão de horas. Tente novamente daqui a pouco ou mude para outro modelo.
Solicitar este modelo no Discord