a2rl.Simulator.gpt_sample_n_steps#

Simulator.gpt_sample_n_steps(seq, n_steps, start_col_index, sample=True)[source]#

Given a GPT token sequence, sample the next n_steps of GPT tokens.

Parameters:

  • seq (ndarray) – This is GPT token sequence as a 1D array.

  • n_steps (int) – Number of next token to predict.

  • start_col_index (int) – Indicate the starting dataframe column index.

Return type:

ndarray

Returns:

Concatenated of original sequence with new predicted sequence.