使用 examples/pytorch/llm/scripts/qwen1half_32b_chat/lora_mp/sft.sh 脚本微调时候的问题

[yezhongxiuchan]

Describe the bug 就用的 该脚本做的 sft， 模型可以正确加载，但是 加载数据的时候出错 “TypeError: Value.init() missing 1 required positional argument: 'dtype'”

报错情况为： sft_main() File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/swift/utils/run_utils.py", line 31, in x_main result = llm_x(args, kwargs) File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/swift/llm/sft.py", line 134, in llm_sft train_dataset, val_dataset = get_dataset( File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/swift/llm/utils/dataset.py", line 1683, in get_dataset dataset = get_function( File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/swift/llm/utils/dataset.py", line 396, in get_dataset_from_repo dataset = load_ms_dataset(dataset_id, subset_split_list) File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/swift/llm/utils/dataset.py", line 252, in load_ms_dataset dataset = MsDataset.load( File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/swift/llm/utils/utils.py", line 99, in _msdataset_ddp_load dataset = _old_msdataset_load(*args, kwargs) File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/modelscope/msdatasets/ms_dataset.py", line 259, in load dataset_context_config).load_dataset( File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/modelscope/msdatasets/data_loader/data_loader_manager.py", line 74, in load_dataset return hf_data_loader( File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/datasets/load.py", line 2523, in load_dataset builder_instance = load_dataset_builder( File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/datasets/load.py", line 2195, in load_dataset_builder dataset_module = dataset_module_factory( File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/datasets/load.py", line 1758, in dataset_module_factory ).get_module() File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/datasets/load.py", line 1071, in get_module { File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/datasets/load.py", line 1072, in config_name: DatasetInfo.from_dict(dataset_info_dict) File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/datasets/info.py", line 356, in from_dict return cls({k: v for k, v in dataset_info_dict.items() if k in field_names}) File "", line 21, in init File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/datasets/info.py", line 174, in post_init self.features = Features.from_dict(self.features) File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/datasets/features/features.py", line 1704, in from_dict obj = generate_from_dict(dic) File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/datasets/features/features.py", line 1359, in generate_from_dict return {key: generate_from_dict(value) for key, value in obj.items()} File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/datasets/features/features.py", line 1359, in return {key: generate_from_dict(value) for key, value in obj.items()} File "/data/anaconda3/envs/cuda12.1/lib/python3.10/site-packages/datasets/features/features.py", line 1367, in generate_from_dict return class_type({k: v for k, v in obj.items() if k in field_names}) TypeError: Value.init() missing 1 required positional argument: 'dtype'

Your hardware and system info Write your system info like CUDA version/system/GPU/torch version here(在这里给出硬件信息和系统信息，如CUDA版本，系统，GPU型号和torch版本等) cuda 12.1 centos7.9 GPU 2*A800 modelscope==1.13.3 torch==2.1.2

Additional context 该环境原来可以可以正常 sft qwen1.5-72b-chat模型的，现在居然也报这个错误了，非常奇怪啊

[wangxingjun778]

尝试升级modelscope到1.14.0