skip exl2 tests on rocm

integration-tests/models/test_flash_awq.py

@@ -9,7 +9,7 @@ def flash_llama_awq_handle(launcher):
         # On ROCm, for awq checkpoints, we need to use gptq kernel that supports ROCm.
         quantize = "gptq"
     elif SYSTEM == "xpu":
-        pytest.skiptest("AWQ is not supported on xpu")
+        pytest.skip("AWQ is not supported on xpu")
     else:
         quantize = "awq"
 

integration-tests/models/test_flash_awq_sharded.py

@@ -4,12 +4,11 @@ from testing_utils import SYSTEM, is_flaky_async
 
 
 @pytest.fixture(scope="module")
+@require_backend_async("cuda", "rocm")
 def flash_llama_awq_handle_sharded(launcher):
     if SYSTEM == "rocm":
         # On ROCm, for awq checkpoints, we need to use gptq kernel that supports ROCm.
         quantize = "gptq"
-    elif SYSTEM == "xpu":
-        pytest.skiptest("AWQ is not supported on xpu")
     else:
         quantize = "awq"
 
@@ -22,6 +21,7 @@ def flash_llama_awq_handle_sharded(launcher):
 
 
 @pytest.fixture(scope="module")
+@require_backend_async("cuda", "rocm")
 async def flash_llama_awq_sharded(flash_llama_awq_handle_sharded):
     await flash_llama_awq_handle_sharded.health(300)
     return flash_llama_awq_handle_sharded.client
@@ -29,6 +29,7 @@ async def flash_llama_awq_sharded(flash_llama_awq_handle_sharded):
 
 @is_flaky_async(max_attempts=5)
 @pytest.mark.asyncio
+@require_backend_async("cuda", "rocm")
 async def test_flash_llama_awq_sharded(flash_llama_awq_sharded, response_snapshot):
     response = await flash_llama_awq_sharded.generate(
         "What is Deep Learning?", max_new_tokens=10, decoder_input_details=True
@@ -47,14 +48,12 @@ async def test_flash_llama_awq_sharded(flash_llama_awq_sharded, response_snapsho
         assert response == response_snapshot
 
 
+@require_backend_async("cuda")
 @pytest.mark.asyncio
 async def test_flash_llama_awq_load_sharded(
     flash_llama_awq_sharded, generate_load, response_snapshot
 ):
-    if SYSTEM == "rocm":
-        pytest.skiptest(
-            "This test relies on ExllamaV2 on ROCm systems, which is highly non-determinstic (flaky)"
-        )
+    # TODO: This test is highly non-deterministic on ROCm.
 
     responses = await generate_load(
         flash_llama_awq_sharded, "What is Deep Learning?", max_new_tokens=10, n=4

integration-tests/models/test_flash_llama_exl2.py

@@ -1,7 +1,9 @@
 import pytest
+from testing_utils import require_backend_async
 
 
 @pytest.fixture(scope="module")
+@require_backend_async("cuda")
 def flash_llama_exl2_handle(launcher):
     with launcher(
         "turboderp/Llama-3-8B-Instruct-exl2",
@@ -16,11 +18,13 @@ def flash_llama_exl2_handle(launcher):
 
 
 @pytest.fixture(scope="module")
+@require_backend_async("cuda")
 async def flash_llama_exl2(flash_llama_exl2_handle):
     await flash_llama_exl2_handle.health(300)
     return flash_llama_exl2_handle.client
 
 
+@require_backend_async("cuda")
 @pytest.mark.asyncio
 @pytest.mark.private
 async def test_flash_llama_exl2(flash_llama_exl2, ignore_logprob_response_snapshot):
@@ -32,6 +36,7 @@ async def test_flash_llama_exl2(flash_llama_exl2, ignore_logprob_response_snapsh
     assert response == ignore_logprob_response_snapshot
 
 
+@require_backend_async("cuda")
 @pytest.mark.asyncio
 @pytest.mark.private
 async def test_flash_llama_exl2_all_params(
@@ -58,6 +63,7 @@ async def test_flash_llama_exl2_all_params(
     assert response == ignore_logprob_response_snapshot
 
 
+@require_backend_async("cuda")
 @pytest.mark.asyncio
 @pytest.mark.private
 async def test_flash_llama_exl2_load(