# LoRA

**URL:** https://discuss.vllm.ai/c/features/lora/16.md

[Latest](https://discuss.vllm.ai/latest.md) · [Categories](https://discuss.vllm.ai/categories.md)

---

## [About the LoRA category](https://discuss.vllm.ai/t/about-the-lora-category/84)

<div class="topic-metadata">

**Author:** [@DarkLight1337](https://discuss.vllm.ai/u/DarkLight1337)\
**Replies:** 0

</div>

---

## [Clarification: Does vLLM support concurrent decoding with multiple LoRA adapters in online inference?](https://discuss.vllm.ai/t/clarification-does-vllm-support-concurrent-decoding-with-multiple-lora-adapters-in-online-inference/1482)

<div class="topic-metadata">

**Author:** [@WarriorOfLiberation](https://discuss.vllm.ai/u/WarriorOfLiberation)\
**Replies:** 1\
**Last updated:** [August 29, 2025, 12:51pm UTC](https://discuss.vllm.ai/t/clarification-does-vllm-support-concurrent-decoding-with-multiple-lora-adapters-in-online-inference/1482 "2025-08-29T12:51:04Z")

</div>

Hi vLLM team, I’m experimenting with multi-LoRA serving in vLLM and I’d like clarification on how batching works in practice. Setup: Started vLLM with --enable-lora --max-lora 2. Loaded two LoRA adapters (say ada…

---

## [RunBot's math-to-text on NVIDIA NeMo Framework AutoModel](https://discuss.vllm.ai/t/runbots-math-to-text-on-nvidia-nemo-framework-automodel/637)

<div class="topic-metadata">

**Author:** [@chris](https://discuss.vllm.ai/u/chris)\
**Replies:** 11\
**Last updated:** [May 19, 2025, 11:32am UTC](https://discuss.vllm.ai/t/runbots-math-to-text-on-nvidia-nemo-framework-automodel/637 "2025-05-19T11:32:11Z")

</div>

NVIDIA spammed me with ~news about their NeMo Framework AutoModel (which seems to related at least somewhat heavily to LoRA, so i picked the LoRA category for this) which claims “Day 0 model support” for popular models (…

---

## [LoRA Adapter enabling with vLLM is not working](https://discuss.vllm.ai/t/lora-adapter-enabling-with-vllm-is-not-working/468)

<div class="topic-metadata">

**Author:** [@uppadas](https://discuss.vllm.ai/u/uppadas)\
**Replies:** 4\
**Last updated:** [April 21, 2025, 8:35pm UTC](https://discuss.vllm.ai/t/lora-adapter-enabling-with-vllm-is-not-working/468 "2025-04-21T20:35:17Z")

</div>

I’m trying to enable lora adapter dynamically with vLLM on AMD MI300x GPU’s. Below are the steps that I followed to enable lora dynamically: \*\*VLLM\_ALLOW\_RUNTIME\_LORA\_UPDATING=True python3 -m vllm.entrypoints.openai.ap…

---

## [Can Lora adapters be loaded on different GPUs](https://discuss.vllm.ai/t/can-lora-adapters-be-loaded-on-different-gpus/397)

<div class="topic-metadata">

**Author:** [@djm-afk](https://discuss.vllm.ai/u/djm-afk)\
**Replies:** 1\
**Last updated:** [April 7, 2025, 3:51pm UTC](https://discuss.vllm.ai/t/can-lora-adapters-be-loaded-on-different-gpus/397 "2025-04-07T15:51:40Z")

</div>

Your current environment Ubuntun How would you like to use vllm I currently need to investigate a question: Does the LoRa adapter for vllm have to run on the same GPU as the base model? For example, if a basic model an…