LoRAForge: Co-Designing Post-Hoc Compression and Execution for Efficient Multi-LoRA Serving
cs.LG, cs.AI
Submitted: 2026-10-31
Updated: 2026-10-31
Code: https://github.com/lishenghui/loraforge
Terminology
Sources
- The Llama 3 Herd of Models
- Compress then Merge: From Multiple LoRAs into One Low-Rank Adapter
- LRAgent: Efficient KV Cache Sharing for Multi-LoRA LLM Agents
- A Rank Stabilization Scaling Factor for Fine-Tuning with LoRA
- Post-Optimization Adaptive Rank Allocation for LoRA
- FraQ: Efficient Coordinate-Space Recompression for Federated Low-Rank Adaptation
- Spectral Transformation for Layer-wise Global Rank Discovery in Federated LoRA for Vision Transformers
- Qwen3 Technical Report
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis