diff --git a/examples/online_serving/kv_events_subscriber.py b/examples/online_serving/kv_events_subscriber.py index 19f6bd5726102..e98ce6c6c9008 100644 --- a/examples/online_serving/kv_events_subscriber.py +++ b/examples/online_serving/kv_events_subscriber.py @@ -28,7 +28,7 @@ class BlockStored(KVCacheEvent): parent_block_hash: ExternalBlockHash | None token_ids: list[int] block_size: int - lora_id: int | None + lora_name: str | None medium: str | None diff --git a/tests/v1/engine/test_engine_core_client.py b/tests/v1/engine/test_engine_core_client.py index 770560a5e549e..09cdc64aba3c1 100644 --- a/tests/v1/engine/test_engine_core_client.py +++ b/tests/v1/engine/test_engine_core_client.py @@ -635,7 +635,7 @@ def test_kv_cache_events( "Block size should be the same as the block size" ) assert event.parent_block_hash is None, "Parent block hash should be None" - assert event.lora_id is None, "Lora id should be None" + assert event.lora_name is None, "Lora name should be None" assert len(event.token_ids) == num_blocks * block_size, ( "Token ids should be the same as the custom tokens" ) diff --git a/tests/v1/kv_connector/unit/test_offloading_connector.py b/tests/v1/kv_connector/unit/test_offloading_connector.py index 69565f584ab89..86d05172e159b 100644 --- a/tests/v1/kv_connector/unit/test_offloading_connector.py +++ b/tests/v1/kv_connector/unit/test_offloading_connector.py @@ -527,7 +527,7 @@ def test_offloading_connector(request_runner): assert event.medium == "A" assert event.token_ids == [] assert event.parent_block_hash is None - assert event.lora_id is None + assert event.lora_name is None event = events[1] assert isinstance(event, BlockRemoved) assert event.block_hashes == to_hashes([4, 5, 6]) diff --git a/vllm/distributed/kv_events.py b/vllm/distributed/kv_events.py index 7b5cb94cf13ea..0795989c11d0e 100644 --- a/vllm/distributed/kv_events.py +++ b/vllm/distributed/kv_events.py @@ -51,7 +51,7 @@ class BlockStored(KVCacheEvent): parent_block_hash: ExternalBlockHash | None token_ids: list[int] block_size: int - lora_id: int | None + lora_name: str | None medium: str | None diff --git a/vllm/distributed/kv_transfer/kv_connector/v1/offloading_connector.py b/vllm/distributed/kv_transfer/kv_connector/v1/offloading_connector.py index 0ad9d4ae1b39f..2243073d61579 100644 --- a/vllm/distributed/kv_transfer/kv_connector/v1/offloading_connector.py +++ b/vllm/distributed/kv_transfer/kv_connector/v1/offloading_connector.py @@ -402,7 +402,7 @@ class OffloadingConnectorScheduler: block_hashes=event.block_hashes, parent_block_hash=None, token_ids=[], - lora_id=None, + lora_name=None, block_size=event.block_size, medium=event.medium, ) diff --git a/vllm/v1/core/block_pool.py b/vllm/v1/core/block_pool.py index 8b0e8fd3a2410..ece9e8dfb2744 100644 --- a/vllm/v1/core/block_pool.py +++ b/vllm/v1/core/block_pool.py @@ -279,7 +279,7 @@ class BlockPool: num_cached_blocks * block_size : num_full_blocks * block_size ], block_size=block_size, - lora_id=request.lora_request.adapter_id + lora_name=request.lora_request.name if request.lora_request else None, medium=MEDIUM_GPU,