Can not train the robot to stand in Training Stage 1

[guangyu0309]

Thanks for your excellent work in advance.
I am running your original code in stage 1 and now the steps just passed 15k. However, at this moment, the standing reward only shows three step-like increase, which is not the same as shown in your paper (in Fig.3 the fourth step-like increase happens around 8k-9k steps roughly). Is there something I should turned? or should I wait a little longer? please advice.
many thanks

[AngryKevin17]

Thank you very much for your kind words and for reporting this issue — and apologies for the delayed response.

We've recently looked into your observation, and it turns out that the discrepancy in the training curve might be related to differences in the Conda environment, to better diagnose the issue, could you please share your current Conda environment configuration? This will help us check for any potential discrepancies in package versions that might affect the training behavior.