[vllm] fix: fix error in vllm patch for diff vllm version and add ci for moe with fp8 rollout (verl-project#4824)

### What does this PR do?

fix error in vllm patch for diff vllm version and add ci for moe with
fp8 rollout

### Checklist Before Starting

- [x] Search for similar PRs. Paste at least one query link here: ...
- [x] Format the PR title as `[{modules}] {type}: {description}` (This
will be checked by the CI)
- `{modules}` include `fsdp`, `megatron`, `sglang`, `vllm`, `rollout`,
`trainer`, `ci`, `training_utils`, `recipe`, `hardware`, `deployment`,
`ray`, `worker`, `single_controller`, `misc`, `perf`, `model`, `algo`,
`env`, `tool`, `ckpt`, `doc`, `data`, `cfg`, `reward`
- If this PR involves multiple modules, separate them with `,` like
`[megatron, fsdp, doc]`
  - `{type}` is in `feat`, `fix`, `refactor`, `chore`, `test`
- If this PR breaks any API (CLI arguments, config, function signature,
etc.), add `[BREAKING]` to the beginning of the title.
  - Example: `[BREAKING][fsdp, megatron] feat: dynamic batching`

### Test

> For changes that can not be tested by CI (e.g., algorithm
implementation, new model support), validate by experiment(s) and show
results like training curve plots, evaluation results, etc.

### API and Usage Example

> Demonstrate how the API changes if any, and provide usage example(s)
if possible.

```python
# Add code snippet or script demonstrating how to use this
```

### Design & Code Changes

> Demonstrate the high-level design if this PR is complex, and list the
specific changes.

### Checklist Before Submitting

> [!IMPORTANT]
> Please check all the following items before requesting a review,
otherwise the reviewer might deprioritize this PR for review.

- [ ] Read the [Contribute
Guide](https://github.com/volcengine/verl/blob/main/CONTRIBUTING.md).
- [ ] Apply [pre-commit
checks](https://github.com/volcengine/verl/blob/main/CONTRIBUTING.md#code-linting-and-formatting):
`pre-commit install && pre-commit run --all-files --show-diff-on-failure
--color=always`
- [ ] Add / Update [the
documentation](https://github.com/volcengine/verl/tree/main/docs).
- [ ] Add unit or end-to-end test(s) to [the CI
workflow](https://github.com/volcengine/verl/tree/main/.github/workflows)
to cover all the code. If not feasible, explain why: ...
- [ ] Once your PR is ready for CI, send a message in [the `ci-request`
channel](https://verl-project.slack.com/archives/C091TCESWB1) in [the
`verl` Slack
workspace](https://join.slack.com/t/verl-project/shared_invite/zt-3855yhg8g-CTkqXu~hKojPCmo7k_yXTQ).
(If not accessible, please try [the Feishu group
(飞书群)](https://applink.larkoffice.com/client/chat/chatter/add_by_link?link_token=772jd4f1-cd91-441e-a820-498c6614126a).)
- [ ] If your PR is related to the `recipe` submodule, please also
update the reference to the submodule commit via `git submodule update
--remote` or `cd recipe && git pull origin main`.

---------

Co-authored-by: Xue Huang <xueh@nvidia.com>

Author: Agoniii

.github/workflows/e2e_ppo_trainer_megatron_vllm_2.yml

@@ -139,6 +139,15 @@ jobs:
           MODEL_ID=Qwen/Qwen3-30B-A3B-Instruct-2507 USE_MBRIDGE=True VANILLA_MBRIDGE=False VALUE_VANILLA_MBRIDGE=False \
           COMMON_PP=2 COMMON_VPP=null COMMON_CP=1 COMMON_TP=4 COMMON_EP=4 COMMON_ETP=1 INFER_TP=8 \
           USE_DIST_CKPT=True ALL_OFFLOAD=True SKIP_SAVE_HF_MODEL=1 bash tests/special_e2e/run_ppo_trainer_megatron.sh
+      - name: Running GSM8K E2E training tests with 3D parallelism with FP8 rollout on 8 L20 GPUs with Megatron-Bridge (Qwen3-30B-A3B-Instruct-2507)
+        run: |
+          ray stop --force
+          ADV_ESTIMATOR=grpo USE_DUMMY_MODEL=True DUMMY_MODEL_CONFIG_PATH=tests/special_e2e/ppo_trainer/expert_parallel/qwen2moe_minimal.json \
+          PPO_MAX_TOKEN_LEN=1024 FWD_MAX_TOKEN_LEN=1024 \
+          MAX_PROMPT_LENGTH=512 MAX_RESPONSE_LENGTH=512 \
+          MODEL_ID=Qwen/Qwen3-30B-A3B-Instruct-2507 USE_MBRIDGE=True VANILLA_MBRIDGE=False VALUE_VANILLA_MBRIDGE=False \
+          COMMON_PP=2 COMMON_VPP=null COMMON_CP=1 COMMON_TP=4 COMMON_EP=4 COMMON_ETP=1 INFER_TP=2 \
+          USE_DIST_CKPT=True ALL_OFFLOAD=True SKIP_SAVE_HF_MODEL=1 ROLLOUT_QUANTIZATION=fp8 bash tests/special_e2e/run_ppo_trainer_megatron.sh
       - name: clean up
         run: |
           rm -rf checkpoints

verl/utils/vllm/vllm_fp8_utils.py

@@ -323,7 +323,10 @@ def _create_param_from_subclass_attributes(custom_param):
 
     del layer.weight_scale_inv
 
-    maybe_post_process_fp8_weight_block(layer)
+    if version.parse(vllm.__version__) == version.parse("0.11.0"):
+        maybe_post_process_fp8_weight_block(layer, self.cutlass_block_fp8_supported)
+    else:
+        maybe_post_process_fp8_weight_block(layer)
 
 
 def process_weights_after_loading_moe_for_vllm10(self, layer) -> None:
@@ -404,7 +407,6 @@ def _create_param_from_subclass_attributes(custom_data, custom_weight):
 
 def process_weights_after_loading_moe_for_vllm11(self, layer) -> None:
     """This function is used to process the weights after loading for a FusedMoE layer, it is used for vllm 0.11"""
-    from vllm.model_executor.layers.fused_moe.rocm_aiter_fused_moe import is_rocm_aiter_moe_enabled
     from vllm.model_executor.layers.quantization.utils.flashinfer_utils import (
         swap_w13_to_w31,
     )
@@ -417,7 +419,14 @@ def process_weights_after_loading_moe_for_vllm11(self, layer) -> None:
         is_deep_gemm_e8m0_used,
     )
 
-    self.rocm_aiter_moe_enabled = is_rocm_aiter_moe_enabled()
+    try:
+        from vllm.model_executor.layers.fused_moe.rocm_aiter_fused_moe import is_rocm_aiter_moe_enabled
+
+        self.rocm_aiter_moe_enabled = is_rocm_aiter_moe_enabled()
+    except ImportError:
+        from vllm._aiter_ops import rocm_aiter_ops
+
+        self.rocm_aiter_moe_enabled = rocm_aiter_ops.is_fused_moe_enabled()
 
     assert self.block_quant and self.quant_config.is_checkpoint_fp8_serialized
     assert self.quant_config.activation_scheme == "dynamic"