Module Owl_rllib_llmSource

Sourcemodule Lm : sig ... end

A toy autoregressive language model.

Sourcemodule Lm_env : sig ... end

Token generation as an ordinary episodic environment (the MDP view), so that the classic algorithms apply directly.

Sourcemodule Lm_rl : sig ... end

RL for language models, following the progression REINFORCE, RLOO, PPO, GRPO and GRPO's variants (Dr. GRPO, DAPO, GSPO, CISPO).

Sourcemodule Lm_task : sig ... end

A verifiable toy task for language-model RL: two-digit addition.