多轮对话问问题之后直接报错
caowenhero opened this issue · comments
多轮对话问问题之后报错
lib/python3.9/site-packages/tensor_parallel/sharding.py:106: UserWarning: operator() sees varying value in profiling, ignoring and this should be handled by GUARD logic (Triggered internally at ../third_party/nvfuser/csrc/parser.cpp:3777.) combined_params = _combine_shards(flat_shards, self._shard_sizes_with_pad, self._sharded_param_shapes) /wxgh/anaconda3/envs/chtglm2/lib/python3.9/site-packages/tensor_parallel/sharding.py:106: UserWarning: operator() profile_node %24 : int[] = prim::profile_ivalue[profile_failed="varying profile values"](%22) does not have profile information (Triggered internally at ../third_party/nvfuser/csrc/graph_fuser.cpp:104.) combined_params = _combine_shards(flat_shards, self._shard_sizes_with_pad, self._sharded_param_shapes)