bugfix: fix the enum/int type mismatch mentioned in #2507 (#2508)

<!-- .github/pull_request_template.md -->

## 📌 Description

As mentioned in #2507, `trtllm_fp8_per_tensor_scale_moe` function would
fail when passed integer `activation_type`.
This PR fixes the type mismatch.

## 🔍 Related Issues

<!-- Link any related issues here -->

## 🚀 Pull Request Checklist

Thank you for contributing to FlashInfer! Before we review your pull
request, please make sure the following items are complete.

### ✅ Pre-commit Checks

- [x] I have installed `pre-commit` by running `pip install pre-commit`
(or used your preferred method).
- [x] I have installed the hooks with `pre-commit install`.
- [x] I have run the hooks manually with `pre-commit run --all-files`
and fixed any reported issues.

> If you are unsure about how to set up `pre-commit`, see [the
pre-commit documentation](https://pre-commit.com/).

## 🧪 Tests

- [x] Tests have been added or updated as needed.
- [ ] All tests are passing (`unittest`, etc.).

## Reviewer Notes

<!-- Optional: anything you'd like reviewers to focus on, concerns, etc.
-->


<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **Refactor**
* Public APIs updated: activation_type now accepts integer values
(defaults adjusted to numeric activation codes).
* Call sites and public function signatures aligned to use the numeric
activation_type.

* **Tests**
* Test inputs updated to supply numeric activation_type values instead
of enum members.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

---------

Signed-off-by: Dimitrios Bariamis <12195802+dbari@users.noreply.github.com>
Co-authored-by: Dimitrios Bariamis <12195802+dbari@users.noreply.github.com>

Author: yzh119

flashinfer/fused_moe/core.py

@@ -989,7 +989,7 @@ def __init__(
             use_deepseek_fp8: bool,
             hidden_size: int,
             intermediate_size: int,
-            activation_type: int = ActivationType.Swiglu,
+            activation_type: int = ActivationType.Swiglu.value,
             use_shuffled_weight: bool = False,
             weight_layout: int = WeightLayout.MajorK,
             use_packed_weights: bool = False,
@@ -1422,7 +1422,7 @@ def trtllm_fp8_per_tensor_scale_moe_op(
         routing_method_type: int = 0,
         enable_pdl: Optional[bool] = None,
         tune_max_num_tokens: int = 8192,
-        activation_type: ActivationType = ActivationType.Swiglu,
+        activation_type: int = ActivationType.Swiglu.value,
     ) -> torch.Tensor:
         if enable_pdl is None:
             enable_pdl = device_support_pdl(hidden_states.device)
@@ -1482,7 +1482,7 @@ def trtllm_fp8_per_tensor_scale_moe_op(
             use_routing_scales_on_input=use_routing_scales_on_input,
             routing_method_type=routing_method_type,
             enable_pdl=enable_pdl,
-            activation_type=activation_type.value,
+            activation_type=activation_type,
         )
         # Call the C++ function
         result = moe_op.trtllm_fp8_per_tensor_scale_moe(
@@ -1507,7 +1507,7 @@ def trtllm_fp8_per_tensor_scale_moe_op(
             routing_method_type,
             enable_pdl,
             [-1, -1] if tactic == -1 else tactic,
-            activation_type.value,
+            activation_type,
         )
 
         return result

tests/moe/test_trtllm_gen_fused_moe.py

@@ -2667,8 +2667,8 @@ def run_moe_test(
 @pytest.mark.parametrize(
     "activation_type",
     [
-        pytest.param(ActivationType.Swiglu, id="Swiglu"),
-        pytest.param(ActivationType.Geglu, id="Geglu"),
+        pytest.param(ActivationType.Swiglu.value, id="Swiglu"),
+        pytest.param(ActivationType.Geglu.value, id="Geglu"),
     ],
 )
 def test_renormalize_routing(
@@ -2855,9 +2855,9 @@ def test_renormalize_routing(
 @pytest.mark.parametrize(
     "activation_type",
     [
-        pytest.param(ActivationType.Swiglu, id="Swiglu"),
-        pytest.param(ActivationType.Geglu, id="Geglu"),
-        pytest.param(ActivationType.Relu2, id="Relu2"),
+        pytest.param(ActivationType.Swiglu.value, id="Swiglu"),
+        pytest.param(ActivationType.Geglu.value, id="Geglu"),
+        pytest.param(ActivationType.Relu2.value, id="Relu2"),
     ],
 )
 def test_deepseekv3_routing(
@@ -2931,8 +2931,8 @@ def test_deepseekv3_routing(
 @pytest.mark.parametrize(
     "activation_type",
     [
-        pytest.param(ActivationType.Swiglu, id="Swiglu"),
-        pytest.param(ActivationType.Geglu, id="Geglu"),
+        pytest.param(ActivationType.Swiglu.value, id="Swiglu"),
+        pytest.param(ActivationType.Geglu.value, id="Geglu"),
     ],
 )
 def test_topk_routing(
@@ -3005,7 +3005,7 @@ def test_topk_routing(
 @pytest.mark.parametrize(
     "activation_type",
     [
-        pytest.param(ActivationType.Swiglu, id="Swiglu"),
+        pytest.param(ActivationType.Swiglu.value, id="Swiglu"),
     ],
 )
 def test_llama4_routing(