你好师兄，我是一个TS小白，我在跑你的STEP代码一直出现这个张量维度不匹配的问题
Traceback (most recent call last):
File "D:\Experiment\Pycode\STEP-github\step\run.py", line 33, in
launch_training(args.cfg, args.gpus)
File "D:\Experiment\Pycode\STEP-github\basicts\launcher.py", line 20, in launch_training
easytorch.launch_training(cfg=cfg, devices=gpus, node_rank=node_rank)
File "D:\APP\Anaconda\envs\TS\lib\site-packages\easytorch\launcher\launcher.py", line 86, in launch_training
train_dist(cfg)
File "D:\APP\Anaconda\envs\TS\lib\site-packages\easytorch\launcher\launcher.py", line 35, in training_func
raise e
File "D:\APP\Anaconda\envs\TS\lib\site-packages\easytorch\launcher\launcher.py", line 31, in training_func
runner.train(cfg)
File "D:\APP\Anaconda\envs\TS\lib\site-packages\easytorch\core\runner.py", line 339, in train
loss = self.train_iters(epoch, iter_index, data)
File "D:\Experiment\Pycode\STEP-github\basicts\runners\base_tsf_runner.py", line 238, in train_iters
forward_return = list(self.forward(data=data, epoch=epoch, iter_num=iter_num, train=True))
File "D:\Experiment\Pycode\STEP-github\step\step_runner\step_runner.py", line 66, in forward
prediction, pred_adj, prior_adj, gsl_coefficient = self.model(history_data=history_data, long_history_data=long_history_data, future_data=None, batch_seen=iter_num, epoch=epoch)
File "D:\APP\Anaconda\envs\TS\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl
return forward_call(*input, **kwargs)
File "D:\Experiment\Pycode\STEP-github\step\step_arch\step.py", line 65, in forward
y_hat = self.backend(short_term_history, hidden_states=hidden_states, sampled_adj=sampled_adj).transpose(1, 2)
File "D:\APP\Anaconda\envs\TS\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl
return forward_call(*input, **kwargs)
File "D:\Experiment\Pycode\STEP-github\step\step_arch\graphwavenet\model.py", line 187, in forward
gate = self.gate_convsi
File "D:\APP\Anaconda\envs\TS\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl
return forward_call(*input, **kwargs)
File "D:\APP\Anaconda\envs\TS\lib\site-packages\torch\nn\modules\conv.py", line 307, in forward
return self._conv_forward(input, self.weight, self.bias)
File "D:\APP\Anaconda\envs\TS\lib\site-packages\torch\nn\modules\conv.py", line 303, in _conv_forward
return F.conv1d(input, weight, bias, self.stride,
RuntimeError: Expected 2D (unbatched) or 3D (batched) input to conv1d, but got input of size: [4, 32, 207, 2017]
我询问GPT也调试了许久，没能解决这个问题，能向您请教以下解决方案么（我只修改了文件路径，模型里代码我没有进行任何操作,您的另外两个模型我都能跑通，只有这个出现了报错，所以我想跟你请教一下）
如果大佬您有空的话，希望可以得到你的回复
这是个兼容性引起的bug，已经在旧的issue中被提及过许多次了。你可以通过将PyTorch降级到1.10及以下来解决这个问题。
STEP copied to clipboard
zezhishao
