> ## Documentation Index
> Fetch the complete documentation index at: https://benchgen.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Merge & Save a Model

> Save, download, or continue training from a LoRA adapter directly, and merge it into a full checkpoint when you actually need one.

A finished training run always gives you a **LoRA adapter**, a small set of weight deltas on top of the base model. **Merging is optional**: it combines the adapter with its base model into a full, standalone checkpoint, useful when you want to deploy it like any other model with no adapter-loading step. You don't have to merge to download the adapter, save it as a usable model, or use it as the starting point for another training run.

## Download the Adapter

Open the completed job under **Train → Jobs**. The **Actions** panel splits into two independent things: the **LoRA Adapter**, ready immediately, and the **Merged Model**, built only if you ask for it.

![The Actions panel on a completed training job, with a Download Adapter (.zip) button under LoRA Adapter, a Merge Model button under Merged Model, and an empty Export History](https://mintcdn.com/benchgen-8fc81371/4FwYeG_Fd1WvnBa5/images/train/merge/01-actions-after-training.jpg?fit=max\&auto=format\&n=4FwYeG_Fd1WvnBa5\&q=85\&s=f1225e8f93c8618cd7003c8f9e3aedf3)

A completed job's Actions panel: Download Adapter under LoRA Adapter, Merge Model under Merged Model, Export History still empty

Nothing here requires the other. Download the adapter, push it to your Models list, or start a new training run from it, all without ever clicking Merge Model.

## Merging into a full checkpoint

Merge when you want a single, self-contained checkpoint with no separate adapter file, for example to hand off, benchmark alongside non-LoRA models, or deploy somewhere that expects one set of weights.

### 1. Click Merge Model

Under **Merged Model**, click **Merge Model**. The panel shows "Merging in progress…", and a new **Export History** entry appears with a **Preparing** badge. This usually takes a couple of minutes.

![The Actions panel with Merging in progress under Merged Model, and one Export History entry showing a Preparing badge](https://mintcdn.com/benchgen-8fc81371/FddC5uLEIMRz8cT0/images/train/merge/02-merging.jpg?fit=max\&auto=format\&n=FddC5uLEIMRz8cT0\&q=85\&s=963f4b1afe98a7619e78a17310538f1c)

A merge running: the Merged Model action shows 'Merging in progress…' and Export History tracks it as Preparing

### 2. Name it and push to BenchGen

When the merge is ready, click **Save on Platform** on the Export History entry. Enter the name it should appear under (for example `Qwen3-0.6B_math`), then click **Push to BenchGen**.

![The Name this model on BenchGen dialog with a text field pre-filled 'Qwen3-0.6B\_math', and Cancel / Push to BenchGen buttons](https://mintcdn.com/benchgen-8fc81371/4FwYeG_Fd1WvnBa5/images/train/merge/03-name-model.jpg?fit=max\&auto=format\&n=4FwYeG_Fd1WvnBa5\&q=85\&s=c94f7db7009cd4618b42b13b9a5042fe)

The naming dialog before pushing a merged model to BenchGen

### 3. Confirm it's saved

The Export History entry updates to **Ready** and **Pushed**, with its own **Download** and **Save on Platform** buttons, download pulls the merged checkpoint to your own computer at any point, whether or not you've pushed it.

![An Export History entry named Qwen3-0.6B\_math, marked Ready and Pushed, with Download and Saved on Platform buttons](https://mintcdn.com/benchgen-8fc81371/4FwYeG_Fd1WvnBa5/images/train/merge/04-saved-on-platform.jpg?fit=max\&auto=format\&n=4FwYeG_Fd1WvnBa5\&q=85\&s=2f4af2565fce92aa301429f3b9629321)

The Export History entry once the merge is Ready and Pushed, with Download and Save on Platform actions

### 4. Use the model

The saved model now appears in **Models** as a deployed, ready endpoint, with its **model name**, **Endpoint URL**, access **Token**, and format (for example `SAFETENSORS`).

![A model card for Qwen3-0.6B\_math, showing Deployed and ready status, its endpoint details, file size, and SAFETENSORS format](https://mintcdn.com/benchgen-8fc81371/FddC5uLEIMRz8cT0/images/train/merge/05-model-ready.jpg?fit=max\&auto=format\&n=FddC5uLEIMRz8cT0\&q=85\&s=dbe12872e536aba7b83f4526edc0991b)

The merged, pushed model's card, deployed and ready with its endpoint details

## Downloading to your own computer

Both artifacts can be pulled down locally at any time, independently of whether you've saved either to your Models list:

| Artifact | Where |
| - | - |
| LoRA adapter | **Download Adapter (.zip)** under LoRA Adapter in the Actions panel |
| Merged checkpoint | **Download** on its Export History entry, once the merge is Ready |

## Using the adapter without merging

> [card] Save it to run in benchmarks
>
> Saving the adapter to your Models list registers it like any other model, so it becomes selectable as a **Trained** model when you evaluate it against a benchmark in Eval, no merge required.

> [card] Continue training from it
>
> Starting a new training job, the **Platform** tab in the Base Model picker lists your previous runs, adapter included. Pick one to build the next round directly on top of it.

> [!NOTE]
>
> The adapter-only "Save on Platform" action is newer than the screenshot above, if your Actions panel already shows a **Save on Platform** button directly under **LoRA Adapter**, use that; otherwise, download the adapter with the button shown and re-upload it as a model the same way you would [upload any model](https://benchgen.com/docs/train/upload-a-model.md).

## Next Steps

> [card] [Fine-tune a Model](https://benchgen.com/docs/train/fine-tune-a-model.md)
>
> Start a new run, optionally continuing from this one's adapter or merged output.

> [card] [Evaluate an inference model](https://benchgen.com/docs/eval/evaluate-a-running-model.md)
>
> Benchmark the saved model against an environment and read the scores.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.