iHore/vLLM
8
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[XPU][CI] Fix ExampleConnector KV cache device selection (#51806)

Browse Source 
Signed-off-by: zhenwei-intel <zhenwei.liu@intel.com>
Co-authored-by: Kunshang Ji <kunshang.ji@intel.com>
This commit is contained in:
liuzhenwei
2026-08-11 23:40:29 +08:00
committed by GitHub
co-authored by Kunshang Ji
parent 5af7c8dad7
commit d648236199
3 changed files with 3 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.buildkite/intel_jobs/test-intel.yaml
+1 -1
View File
@@ -105,7 +105,7 @@ steps:
pytest -v -s v1/test_serial_utils.py &&
VLLM_DISABLE_COMPILE_CACHE=1 pytest -v -s v1/e2e/general/test_correctness_sliding_window.py &&
pytest -v -s v1/spec_decode --ignore=v1/spec_decode/test_max_len.py --ignore=v1/spec_decode/test_speculators_eagle3.py --ignore=v1/spec_decode/test_acceptance_length.py --ignore=v1/spec_decode/test_speculators_correctness.py &&
pytest -v -s v1/kv_connector/unit --ignore=v1/kv_connector/unit/test_example_connector.py --ignore=v1/kv_connector/unit/test_lmcache_integration.py --ignore=v1/kv_connector/unit/test_hf3fs_client.py --ignore=v1/kv_connector/unit/test_hf3fs_connector.py --ignore=v1/kv_connector/unit/test_hf3fs_metadata_server.py --ignore=v1/kv_connector/unit/test_offloading_connector.py'
pytest -v -s v1/kv_connector/unit --ignore=v1/kv_connector/unit/test_lmcache_integration.py --ignore=v1/kv_connector/unit/test_hf3fs_client.py --ignore=v1/kv_connector/unit/test_hf3fs_connector.py --ignore=v1/kv_connector/unit/test_hf3fs_metadata_server.py --ignore=v1/kv_connector/unit/test_offloading_connector.py'
- label: "XPU server test"
depends_on:
- image-build-xpu
tests/v1/kv_connector/unit/test_example_connector.py
+1 -1
View File
@@ -115,7 +115,7 @@ def process_prompt(processor, llm: LLM, question: str, image_urls: list[Image]):
["FLASH_ATTN", "TRITON_ATTN"]
if current_platform.is_cuda()
else ["TRITON_ATTN"]
if current_platform.is_rocm()
if current_platform.is_rocm() or current_platform.is_xpu()
else []
),
)
vllm/distributed/kv_transfer/kv_connector/v1/example_connector.py
+1 -1
View File
@@ -180,7 +180,7 @@ class ExampleConnector(KVConnectorBase_V1):
layer_name, request.token_ids, request.mm_hashes
)
kv_cache_cpu = safetensors.torch.load_file(filename)["kv_cache"]
kv_cache = kv_cache_cpu.to("cuda", non_blocking=True)
kv_cache = kv_cache_cpu.to(kv_cache_layer.device, non_blocking=True)
if isinstance(attn_metadata, dict):
inject_kv_into_layer(
kv_cache_layer,