Description
您好，复现论文中的数据时遇到问题，训练步骤如下：
1、python main.py --device_id 0 --machine machine --dataset "Crowd*Cellular*BikeNYC2*TaxiNYC2*TrafficCD*TrafficJN*TrafficNJ*TrafficSH*TrafficTJ*TrafficZZ*TrafficHZ" --task short --size middle --mask_strategy_random 'batch' --lr 3e-4 --used_data 'diverse' --prompt_ST 0 --few_ratio 1.0
2、python main.py --device_id 0 --machine machine --task short --size middle --prompt_ST 1 --pred_len 6 --his_len 6 --num_memory_spatial 512 --num_memory_temporal 512 --prompt_content 's_p_c' --dataset "Crowd*Cellular*BikeNYC2*TaxiNYC2*TrafficCD*TrafficJN*TrafficNJ*TrafficSH*TrafficTJ*TrafficZZ*TrafficHZ" --lr 3e-4 --used_data 'diverse' --file_load_path "/home/synebula/文档/UniST/UniST/src/experiments/Pretrain_Dataset_Crowd*Cellular*BikeNYC2*TaxiNYC2*TrafficCD*TrafficJN*TrafficNJ*TrafficSH*TrafficTJ*TrafficZZ*TrafficHZ_Task_short_FewRatio_1.0/model_save/model_best"
3、python main.py --device_id 0 --machine machine --task short --size middle --prompt_ST 1 --pred_len 6 --his_len 6 --num_memory_spatial 512 --num_memory_temporal 512 --prompt_content 's_p_c' --dataset Crowd --used_data 'diverse' --file_load_path "/home/synebula/文档/UniST/UniST/src/experiments/Prompt_Dataset_Crowd*Cellular*BikeNYC2*TaxiNYC2*TrafficCD*TrafficJN*TrafficNJ*TrafficSH*TrafficTJ*TrafficZZ*TrafficHZ_Task_short_FewRatio_0.5/model_save/model_best" --few_ratio 0.0
第三步测试结果是
min: 0.0 max: 593118.0
model initialized memory
model initialized memory
model initialized!
/home/synebula/文档/UniST/UniST/src/main.py:130: FutureWarning: You are using torch.load with weights_only=False (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blomain/SECURITY.md#untrusted-models for more details). In a future release, the default value for weights_only will be flipped to True. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via torch.serialization.add_safe_globals. We recommend you start setting weights_only=True for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
model.load_state_dict(torch.load('{}.pkl'.format(args.file_load_path),map_location=device), strict=False)
pretrained model loaded
RMSE_best:0.00030013404934801994
{'Crowd': {'temporal': {0.5: {'rmse': 5136.990826869795, 'mae': 2341.6112440385614}}}}
是我运行的命令有问题吗