[TRTLLM-5863][feat] Support MoE INT8 Weight-Only-Quantization in PyTorch Workflow#6629

yuxianq merged 2 commits into

NVIDIA:mainNVIDIA/TensorRT-LLM:mainfrom

Yuening-wa:user/yueningl/moe_int8_weight_only_quant_supportYuening-wa/TensorRT-LLM:user/yueningl/moe_int8_weight_only_quant_supportCopy head branch name to clipboard