[Fix] optimize/shard model in TPInferEngine init#4684

tiandiao123 merged 3 commits intohpcaitech:feature/colossal-inferencehpcaitech/ColossalAI:feature/colossal-inferencefrom

yuanheng-zhao:colossal-inference/fix-pryuanheng-zhao/ColossalAI:colossal-inference/fix-prCopy head branch name to clipboard