[求助]无法加载本地的Vicuna-7B模型

[Manticore425]

我尝试使用Vicuna-7B模型，已经把融合过LLaMA原模型的Vicuna模型放入models文件夹中。但是在Web UI中加载时报错：

Loading checkpoint shards: 100%|█████████████████████████████████████████████████████████████████████████████| 2/2 [00:43<00:00, 21.97s/it]
[2023-04-29 23:37:58,280] [INFO] [comm.py:634:init_distributed] Not using the DeepSpeed or dist launchers, attempting to detect MPI environment...
2023-04-29 23:37:58,281 [ERROR] [models.py:595] No module named 'mpi4py'

请问我该怎么解决它？

[Manticore425]

我安装了mpi4py库后，再次加载依然报错：

Loading checkpoint shards: 100%|█████████████████████████████████████████████████████████████████████████████| 2/2 [00:35<00:00, 17.74s/it]
[2023-04-29 23:54:25,667] [INFO] [comm.py:634:init_distributed] Not using the DeepSpeed or dist launchers, attempting to detect MPI environment...
[2023-04-29 23:54:25,739] [INFO] [comm.py:688:mpi_discovery] Discovered MPI settings of world_rank=0, local_rank=0, world_size=1, master_addr=172.26.194.56, master_port=29500
[2023-04-29 23:54:25,739] [INFO] [comm.py:652:init_distributed] Initializing TorchBackend in DeepSpeed with backend nccl
2023-04-29 23:54:25,744 [INFO] [distributed_c10d.py:432] Added key: store_based_barrier_key:1 to store for rank: 0
2023-04-29 23:54:25,745 [INFO] [distributed_c10d.py:466] Rank 0: Completed store-based barrier for key:store_based_barrier_key:1 with 1 nodes.
[2023-04-29 23:54:25,745] [INFO] [logging.py:93:log_dist] [Rank 0] DeepSpeed info: version=0.8.3, git-hash=unknown, git-branch=unknown
2023-04-29 23:54:25,802 [ERROR] [models.py:595] Cannot copy out of meta tensor; no data!