Attempt to run in the dry mode throws an error

[snitovets]

Unable to analyze 
candle (OHLCV) data for pair BTC/USDT:USDT: Weights only load failed. This file can  
still be loaded, to do so you have two options, do those steps only if you trust     
the source of the checkpoint. 
        (1) In PyTorch 2.6, we changed the default value of the `weights_only`       
argument in `torch.load` from `False` to `True`. Re-running `torch.load` with        
`weights_only` set to `False` will likely succeed, but it can result in arbitrary    
code execution. Do it only if you got the file from a trusted source.
        (2) Alternatively, to load with `weights_only=True` please check the
recommended steps in the following error message.
        WeightsUnpickler error: Unsupported global: GLOBAL
freqtrade.freqai.torch.PyTorchModelTrainer.PyTorchLSTMTrainer was not an allowed     
global by default. Please use
`torch.serialization.add_safe_globals([PyTorchLSTMTrainer])` or the
`torch.serialization.safe_globals([PyTorchLSTMTrainer])` context manager to allowlistthis global if you trust this class/function.


Check the documentation of torch.load to learn more about types accepted by default  
with weights_only https://pytorch.org/docs/stable/generated/torch.load.html.

[989888]

My direct run command:freqtrade trade --config user_data/test6_lstm.json --strategy test6_lstm --freqaimodel PyTorchLSTMRegressor --db-url sqlite:///test6_lstm.dry_run.sqlite

[goranv90]

I have the same issue which I cannot solve. Did you manage to solve it?

[Supreet197]

Go to class PyTorchLSTMRegressor
In init add line
super().init(**kwargs)
torch.serialization.add_safe_globals(
[PyTorchLSTMModel, ModuleList, LSTM, BatchNorm1d, Dropout, Linear, AlphaDropout, ReLU,AdamW,defaultdict,dict,DefaultPyTorchDataConvertor,MSELoss,BaseTensorboardLogger,ReduceLROnPlateau])

[rogue2021]

I tried to insert torch.serialization.add_safe_globals(
[PyTorchLSTMModel, ModuleList, LSTM, BatchNorm1d, Dropout, Linear, AlphaDropout, ReLU,AdamW,defaultdict,dict,DefaultPyTorchDataConvertor,MSELoss,BaseTensorboardLogger,ReduceLROnPlateau]) but it was not fixed the issue. I am using the latest Freqtrade 2025.8 stable as well. Please help

[manigot]

@989888 @goranv90 @rogue2021 on config file add "initial_state": "running" .