!!! tip "Shortcut"
You can use the shortcut `agl.VERL(...)` to create a VERL instance.
```python
import agentlightning as agl
agl.VERL(...)
```
pip install agentlightning[verl]!!! warning
To avoid various compatibility issues, follow the steps in the [installation guide](../tutorials/installation.md) to set up VERL and its dependencies. Installing VERL directly with `pip install agentlightning[verl]` can cause issues unless you already have a compatible version of PyTorch installed.
!!! note "Notes for Readers"
[VERL][agentlightning.algorithm.verl.VERL] in this article refers to a wrapper, provided by Agent-lightning, of the [VERL framework](https://github.com/volcengine/verl). It's a subclass of [agentlightning.Algorithm][]. To differentiate it from the VERL framework, all references to the VERL framework shall use the term "VERL framework", and all references to the Agent-lightning wrapper shall be highlighted with a link.
[VERL][agentlightning.algorithm.verl.VERL] expects no initial resources. The first LLM endpoint is directly deployed from the VERL configuration (.actor_rollout_ref.model.path). The resource key is always main_llm.
[VERL][agentlightning.algorithm.verl.VERL] currently does not support optimizing multiple [LLM][agentlightning.LLM]s together.
!!! note
The resource type created by [VERL][agentlightning.algorithm.verl.VERL] is actually a [ProxyLLM][agentlightning.ProxyLLM], a subclass of the [LLM][agentlightning.LLM] type. This object contains a **URL template** provided by [VERL][agentlightning.algorithm.verl.VERL], with placeholders for rollout and attempt IDs. When a rollout begins on the agent side, the framework uses the current `rollout_id` and `attempt_id` to format this template, generating a final, unique endpoint URL. This URL points to [VERL][agentlightning.algorithm.verl.VERL]'s internal proxy, allowing it to intercept and log all traffic for that specific attempt, for tracing and load balancing purposes. For agents created with the `@rollout` decorator, this resolution of the template is handled automatically ("auto-stripped"). Class-based agents will need to manually resolve the [`ProxyLLM`][agentlightning.ProxyLLM] using the rollout context.
```python
proxy_llm = resources["main_llm"]
proxy_llm.get_base_url(rollout.rollout_id, rollout.attempt.attempt_id)
```
Internally, [VERL][agentlightning.algorithm.verl.VERL] decomposes each agent execution into prompt–response pairs via the [Adapter][agentlightning.Adapter] and associates them with their corresponding reward signals as [Triplet][agentlightning.Triplet] objects. The final scalar reward, derived from the last triplet in the trajectory, is propagated to all preceding triplets following the identical assignment strategy. This ensures that each triplet receives an identical reward signal and can be independently optimized as a valid RLHF trajectory within the VERL framework.
At present, [VERL][agentlightning.algorithm.verl.VERL] does not expose fine-grained control over its reward propagation or credit assignment mechanisms. Users requiring customized reward shaping or trajectory decomposition are advised to clone and modify the [VERL][agentlightning.algorithm.verl.VERL] source implementation directly.
- Train SQL Agent with RL - A practical example of training a SQL agent using VERL.
::: agentlightning.algorithm.verl
::: agentlightning.verl