Add Libra to LLM post-training repositories

README.md

@@ -274,6 +274,7 @@ A taxonomy of post-training approaches for **LLMs**, categorized into Fine-tunin
 | 13 | [**trlX**](https://github.com/CarperAI/trlx)                                                               | A distributed training framework for fine-tuning large language models (LLMs) with reinforcement learning. Supports both Accelerate and NVIDIA NeMo backends, allowing training of models up to 20B+ parameters. Implements PPO and ILQL, and integrates with CHEESE for human-in-the-loop data collection.                                                                                                                                                                                              |
 | 14 | [**Okapi**](https://github.com/nlp-uoregon/Okapi)                                                          | A framework for instruction tuning in LLMs with RLHF, supporting 26 languages. Provides multilingual resources such as ChatGPT prompts, instruction datasets, and response ranking data, along with both BLOOM-based and LLaMa-based models and evaluation benchmarks.                                                                                                                                                                                                                                                 |
 | 15 | [**LLaMA-Factory**](https://github.com/hiyouga/LLaMA-Factory)                                              | *Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)*. Supports a wide array of models (e.g., LLaMA, LLaVA, Qwen, Mistral) with methods including pre-training, multimodal fine-tuning, reward modeling, PPO, DPO, and ORPO. Offers scalable tuning (16-bit, LoRA, QLoRA) with advanced optimizations and logging integrations, and provides fast inference via API, Gradio UI, and CLI with vLLM workers.                                                 |
+| 16 | [**Libra**](https://github.com/NetX-lab/Libra)                                                               | A resource-aware systems framework for disaggregated and asynchronous post-training of agentic language models. It coordinates training and rollout pools, supports heterogeneous vLLM workers, dynamic resource planning, elastic execution, and asynchronous GRPO with agentic workloads. [Paper](https://arxiv.org/abs/2606.03077) |
 
 ---
 ## ⚡ Applications & Benchmarks