[Bugfix][Speculative Decoding] Fix Eagle3 quantization config issue (#25883)
Signed-off-by: Rahul Tuli <rtuli@redhat.com>
This commit is contained in:
@@ -14,6 +14,9 @@ from vllm.model_executor.models.interfaces import supports_eagle3
|
||||
pytest.param(
|
||||
"nm-testing/Speculator-Qwen3-8B-Eagle3-converted-071-quantized",
|
||||
id="qwen3-eagle3-speculator"),
|
||||
pytest.param(
|
||||
"nm-testing/Speculator-Qwen3-8B-Eagle3-converted-071-quantized-w4a16",
|
||||
id="qwen3-eagle3-speculator-w4a16-verifier"),
|
||||
])
|
||||
def test_eagle3_speculators_model(vllm_runner, example_prompts, model_path,
|
||||
monkeypatch):
|
||||
|
||||
Reference in New Issue
Block a user