What is the model suitable for time series forecasting?

I would like to forecast a time series of commodity prices.
At the moment, I am using LSTM, CNN, and transformer, but the accuracy is not improving.

Hi @ruorch

The accuracy of the model also depends on the how well the dataset preprocessed, what are the optimizer components has been used in the model and model trained for how many epochs. Please refer to this timeseries forecasting model for your reference.

Please provide us the reproducible code to replicate the error and fix the issue. Thank you.

@ruorch On the business side – My first advice would be to NOT expect too much from any commodity price forecasting model. Anytime you come across a commodity price forecast, be sure someone is selling something.
That being said what have you tried so far more specifically? What data frequency are we talking about here? What feature engineering have you done? Do you integrate external factors (SnD balances, weather, global macro, CFTC/money flow, energy prices, exchange rates, …)? What forecasting horizon? Basically what commodity are you interested in? How far are you from what you expected? What kind of model do you use as a benchmark for you predictions?
Thank you