回测时加载模型checkpoint和初始化actor critic维度不同

[wuxiawei]

在执行：
results = DRLAgent.DRL_prediction_load_from_file(model_name='maesac',environment=test_trade_gym, cwd=model_path)
的时候报错：
RuntimeError: Error(s) in loading state_dict for SACPolicy:
size mismatch for actor.mu.weight: copying a param with shape torch.Size([88, 256]) from checkpoint, the shape in current model is torch.Size([31064, 256]).
size mismatch for actor.mu.bias: copying a param with shape torch.Size([88]) from checkpoint, the shape in current model is torch.Size([31064]).
size mismatch for actor.log_std.weight: copying a param with shape torch.Size([88, 256]) from checkpoint, the shape in current model is torch.Size([31064, 256]).
size mismatch for actor.log_std.bias: copying a param with shape torch.Size([88]) from checkpoint, the shape in current model is torch.Size([31064]).
size mismatch for critic.qf0.0.weight: copying a param with shape torch.Size([256, 11440]) from checkpoint, the shape in current model is torch.Size([256, 42416]).
size mismatch for critic.qf1.0.weight: copying a param with shape torch.Size([256, 11440]) from checkpoint, the shape in current model is torch.Size([256, 42416]).
size mismatch for critic_target.qf0.0.weight: copying a param with shape torch.Size([256, 11440]) from checkpoint, the shape in current model is torch.Size([256, 42416]).
size mismatch for critic_target.qf1.0.weight: copying a param with shape torch.Size([256, 11440]) from checkpoint, the shape in current model is torch.Size([256, 42416]).
请问这是什么原因呢？

[xbkaishui]

请问这个问题解决了吗？

[wuxiawei]

还没有，你现在解决了吗？

[yo-yoo]

请问这个问题解决了吗？

遇到了同样的问题，请问解决了吗？是因为单卡运行导致的吗？

[mzliu2017]

检查了好久，终于找到出错是在zip存取和读取数据有错，存储时data_to_json()之前的data格式正确，为例
，后面读取时json_to_data()会产生warning，导致读取错误。

修改numpy版本可以解决此bug，我之前使用的是numpy=1.24.0，改为1.23.5就可以使用了（python=3.8.19），但是更底层的原因就没有深究了，希望对你有帮助~