Boreas813

I modified the function to reduce the hardware requirements to 2.3GB and a 20-minute conversion process, have a try.

> As of pandas 2.0, append (previously deprecated) [was removed](https://pandas.pydata.org/docs/whatsnew/v2.0.0.html#removal-of-prior-version-deprecations-changes). > > You need to use [concat](https://pandas.pydata.org/docs/reference/api/pandas.concat.html) instead (for most applications): > > df_account_value = pd.concat([df_account_value, temp], ignore_index=True) > >...

> Is the algorithm running for 2731 seconds including data import, preprocessing, and training, or just the training time? Both times are the time it takes to run all the...

> > > Is the algorithm running for 2731 seconds including data import, preprocessing, and training, or just the training time? > > > > > > Both times are...

It could be batch size too small? I don't see any doc in sb3 readthedocs that introduce how to train on gpu.

> > It could be batch size too small? I don't see any doc in sb3 readthedocs that introduce how to train on gpu. > > ooh, SB3 is not...