# Symbiosis: Multi-Adapter Inference and Fine-Tuning

arXiv: https://arxiv.org/abs/2507.03220 · HTML: https://mlsanity.com/paper/2507.03220/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 29 of its 59 references that are in the map.
- First posted on arXiv 3 Jul 2025.

## Builds on (most cited first, 29 of 29)

- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Few-Shot Parameter-Efficient Fine-Tuning is Better and Cheaper than In-Context Learning](https://mlsanity.com/paper/2205.05638/) (2022)
- [PyTorch FSDP: Experiences on Scaling Fully Sharded Data Parallel](https://mlsanity.com/paper/2304.11277/) (2023)
- [Stealing Machine Learning Models via Prediction APIs](https://mlsanity.com/paper/1609.02943/) (2016)
- [DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving](https://mlsanity.com/paper/2401.09670/) (2024)
- [FlexGen: High-Throughput Generative Inference of Large Language Models with a Single GPU](https://mlsanity.com/paper/2303.06865/) (2023)
- [Splitwise: Efficient generative LLM inference using phase splitting](https://mlsanity.com/paper/2311.18677/) (2023)
- [DeepSpeed Inference: Enabling Efficient Inference of Transformer Models at Unprecedented Scale](https://mlsanity.com/paper/2207.00032/) (2022)
- [DeepSpeed-MoE: Advancing Mixture-of-Experts Inference and Training to Power Next-Generation AI Scale](https://mlsanity.com/paper/2201.05596/) (2022)
- [DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Heads](https://mlsanity.com/paper/2410.10819/) (2024)
- [ZeRO-Infinity: Breaking the GPU Memory Wall for Extreme Scale Deep Learning](https://mlsanity.com/paper/2104.07857/) (2021)
- [InfiniGen: Efficient Generative Inference of Large Language Models with Dynamic KV Cache Management](https://mlsanity.com/paper/2406.19707/) (2024)
- [SARATHI: Efficient LLM Inference by Piggybacking Decodes with Chunked Prefills](https://mlsanity.com/paper/2308.16369/) (2023)
- [Scaling Down to Scale Up: A Guide to Parameter-Efficient Fine-Tuning](https://mlsanity.com/paper/2303.15647/) (2023)
- [Cost-Efficient Large Language Model Serving for Multi-turn Conversations with CachedAttention](https://mlsanity.com/paper/2403.19708/) (2024)
- [S-LoRA: Serving Thousands of Concurrent LoRA Adapters](https://mlsanity.com/paper/2311.03285/) (2023)
- [DeepSpeed-FastGen: High-throughput Text Generation for LLMs via MII and DeepSpeed-Inference](https://mlsanity.com/paper/2401.08671/) (2024)
- [Efficient LLM Scheduling by Learning to Rank](https://mlsanity.com/paper/2408.15792/) (2024)
- [Chimera: Efficiently Training Large-Scale Neural Networks with Bidirectional Pipelines](https://mlsanity.com/paper/2107.06925/) (2021)
- [RoSA: Accurate Parameter-Efficient Fine-Tuning via Robust Adaptation](https://mlsanity.com/paper/2401.04679/) (2024)
- [Training Large Neural Networks with Constant Memory using a New Execution Algorithm](https://mlsanity.com/paper/2002.05645/) (2020)
- [CaraServe: CPU-Assisted and Rank-Aware LoRA Serving for Generative LLM Inference](https://mlsanity.com/paper/2401.11240/) (2024)
- [Model Extraction Attacks Revisited](https://mlsanity.com/paper/2312.05386/) (2023)
- [InstInfer: In-Storage Attention Offloading for Cost-Effective Long-Context LLM Inference](https://mlsanity.com/paper/2409.04992/) (2024)
- [LayerKV: Optimizing Large Language Model Serving with Layer-wise KV Cache Management](https://mlsanity.com/paper/2410.00428/) (2024)
- [Chameleon: Adaptive Caching and Scheduling for Many-Adapter LLM Inference Environments](https://mlsanity.com/paper/2411.17741/) (2024)
- [HoneypotNet: Backdoor Attacks Against Model Extraction](https://mlsanity.com/paper/2501.01090/) (2025)
- [SYMPHONY: Improving Memory Management for LLM Inference Workloads](https://mlsanity.com/paper/2412.16434/) (2024)

## Abstract (arXiv)

Parameter-efficient fine-tuning (PEFT) allows model builders to capture the task-specific parameters into adapters, which are a fraction of the size of the original base model. Popularity of PEFT technique for fine-tuning has led to the creation of a large number of adapters for popular Large Language Models (LLMs). However, existing frameworks fall short in supporting inference or fine-tuning with multiple adapters in the following ways. 1) For fine-tuning, each job needs to deploy its dedicated base model instance, which results in excessive GPU memory consumption and poor GPU utilization. 2) While popular inference platforms can serve multiple PEFT adapters, they do not allow independent resource management or mixing of different PEFT methods. 3) They cannot make effective use of heterogeneous accelerators. 4) They do not provide privacy to users who may not wish to expose their fine-tuned parameters to service providers. In Symbiosis, we address the above problems by enabling the as-a-service deployment of the base model. The base model layers can be shared across multiple inference or fine-tuning processes. Our split-execution technique decouples the execution of client-specific adapters and layers from the frozen base model layers offering them flexibility to manage their resources, to select their fine-tuning method, to achieve their performance goals. Our approach is transparent to models and works out-of-the-box for most models in the transformers library. We demonstrate the use of Symbiosis to simultaneously fine-tune 20 Gemma2-27B adapters on 8 GPUs.