RL / post-training libraries
RL / post-training libraries covers RLHF, DPO and RL trainer libraries (verl, OpenRLHF, TRL's RL trainers).
In brief
RL / post-training libraries, in brief.
What it isRL / post-training libraries covers RLHF, DPO and RL trainer libraries (verl, OpenRLHF, TRL's RL trainers).
Where it sitsOne of 12 categories in the Training and fine-tuning layer. Training and fine-tuning covers the frameworks, libraries and platforms used to train a model from scratch or adapt an existing one to a task, and the tools that track those experiments.
What we track20 products, including Agent Lightning, AReaL, ART (Agent Reinforcement Trainer), Axolotl and DeepSpeed-Chat.
The products
20 products in RL / post-training libraries.
Each name opens a search for that product in our index.
ProductMakerIts own site
From our census of 2026-11-08. A product listed here is one that belongs in this category; the search answer says whether we have checked it.
Questions
RL / post-training libraries, answered.
What is RL / post-training libraries in AI infrastructure?RL / post-training libraries covers RLHF, DPO and RL trainer libraries (verl, OpenRLHF, TRL's RL trainers).
Which layer of AI infrastructure is RL / post-training libraries in?RL / post-training libraries is in the Training and fine-tuning layer. Training and fine-tuning covers the frameworks, libraries and platforms used to train a model from scratch or adapt an existing one to a task, and the tools that track those experiments.
Which products are in RL / post-training libraries?We track 20 products in RL / post-training libraries, including Agent Lightning, AReaL, ART (Agent Reinforcement Trainer), Axolotl, DeepSpeed-Chat, LLaMA-Factory, ms-swift and NeMo RL (successor to NeMo-Aligner). The full list is on this page.
Training and fine-tuning
The other categories in Training and fine-tuning.
AutoML & hyperparameter optimizationComputer vision toolkits & platformsDeep learning & ML frameworksDistributed training frameworksExperiment trackingFeature storesFederated & privacy-preserving learningFine-tuning platformsML workflow / MLOps platformsOpen-source fine-tuning librariesRL environments & gyms