From a5d19cbb95872c4b426c06735733568542fa33db Mon Sep 17 00:00:00 2001 From: Nick Hill Date: Wed, 8 Jul 2026 19:30:11 +0100 Subject: [PATCH] [Core] Move MRV1 `late_interaction_runner.py` out of MRV2 subtree (#48014) Signed-off-by: Nick Hill --- tests/v1/worker/test_late_interaction_runner.py | 2 +- vllm/v1/{worker/gpu => }/pool/late_interaction_runner.py | 0 vllm/v1/worker/gpu_model_runner.py | 2 +- 3 files changed, 2 insertions(+), 2 deletions(-) rename vllm/v1/{worker/gpu => }/pool/late_interaction_runner.py (100%) diff --git a/tests/v1/worker/test_late_interaction_runner.py b/tests/v1/worker/test_late_interaction_runner.py index 9719485cd54..5f1b7cff37c 100644 --- a/tests/v1/worker/test_late_interaction_runner.py +++ b/tests/v1/worker/test_late_interaction_runner.py @@ -11,7 +11,7 @@ from vllm.v1.pool.late_interaction import ( build_late_interaction_doc_params, build_late_interaction_query_params, ) -from vllm.v1.worker.gpu.pool.late_interaction_runner import LateInteractionRunner +from vllm.v1.pool.late_interaction_runner import LateInteractionRunner def _make_pooling_params( diff --git a/vllm/v1/worker/gpu/pool/late_interaction_runner.py b/vllm/v1/pool/late_interaction_runner.py similarity index 100% rename from vllm/v1/worker/gpu/pool/late_interaction_runner.py rename to vllm/v1/pool/late_interaction_runner.py diff --git a/vllm/v1/worker/gpu_model_runner.py b/vllm/v1/worker/gpu_model_runner.py index 199470aaf87..dd214a6f3e8 100644 --- a/vllm/v1/worker/gpu_model_runner.py +++ b/vllm/v1/worker/gpu_model_runner.py @@ -174,6 +174,7 @@ from vllm.v1.outputs import ( SamplerOutput, make_empty_encoder_model_runner_output, ) +from vllm.v1.pool.late_interaction_runner import LateInteractionRunner from vllm.v1.pool.metadata import PoolingMetadata, PoolingStates from vllm.v1.sample.logits_processor import LogitsProcessors, build_logitsprocs from vllm.v1.sample.logits_processor.interface import LogitsProcessor @@ -207,7 +208,6 @@ from vllm.v1.worker.cp_utils import ( from vllm.v1.worker.dp_utils import coordinate_batch_across_dp from vllm.v1.worker.ec_connector_model_runner_mixin import ECConnectorModelRunnerMixin from vllm.v1.worker.gpu.attn_utils import _reshape_attention_kv_cache -from vllm.v1.worker.gpu.pool.late_interaction_runner import LateInteractionRunner from vllm.v1.worker.gpu_input_batch import CachedRequestState, InputBatch from vllm.v1.worker.gpu_ubatch_wrapper import UBatchWrapper from vllm.v1.worker.kv_connector_model_runner_mixin import KVConnectorModelRunnerMixin