试了一下nv build的minimax2.7速度还行,平均响应时间30s。
所以还是要挑不火的模型用。。像是glm还是避开吧。。。
https://build.nvidia.com/models?filters=publisher%3Amoonshotai%2Cpublisher%3Aminimaxai%2Cpublisher%3Az_ai&orderBy=weightPopular%3ADESC
所以还是要挑不火的模型用。。像是glm还是避开吧。。。
https://build.nvidia.com/models?filters=publisher%3Amoonshotai%2Cpublisher%3Aminimaxai%2Cpublisher%3Az_ai&orderBy=weightPopular%3ADESC