[Pooling] Report LoRA adapter names in responses (#56004)
Signed-off-by: Taneem Ibrahim <taneem.ibrahim@gmail.com>
Taneem Ibrahim committed
Sep 9, 2026 at 12:31 UTC
af51d0ef03ecf62302b51e7ab3042b088449d6b5
2 files changed
+2
-2
vllm/entrypoints/pooling/base/serving.py
+1
-1
@@ -108,7 +108,6 @@ class PoolingBaseServing(ABC, BaseServing):
108
request: AnyPoolingRequest,
109
raw_request: Request | None = None,
110
):
111
- model_name = self.models.model_name()
111
base_request_id = self._base_request_id(
112
raw_request, getattr(request, "request_id", None)
113
)
@@ -117,6 +116,7 @@ class PoolingBaseServing(ABC, BaseServing):
116
117
pooling_params = io_processor.create_pooling_params(request)
118
lora_request = self._maybe_get_adapters(request)
119
+ model_name = self.models.model_name(lora_request)
120
priorities = getattr(request, "priority", 0)
121
prompt_extras = {
122
k: v
vllm/entrypoints/pooling/scoring/serving.py
+1
-1
@@ -79,7 +79,7 @@ class ServingScores(PoolingServing):
79
final_res_batch = ctx.final_res_batch
80
request_id = ctx.request_id
81
created_time = ctx.created_time
82
- model_name = self.models.model_name()
82
+ model_name = ctx.model_name
83
84
if isinstance(ctx.request, ScoreRequest):
85
return self._request_output_to_score_response(