need action map function for language table dataset

hello,
thanks for your great work.
I'm trying to train the RT-1-X model and evaluate it on different benchmarks like language table. I wonder how do I map actions generated by the model and actions needed by the benchmarks? I'm expecting the action map function in the colab notebooks that you have provided, but the function for language table is still TODO.
also is there other methods to evaluate the model?