biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Core] Remove prompt string from engine core data structures (#17214)

Browse Source 
Signed-off-by: Nick Hill <nhill@redhat.com>
This commit is contained in:
Nick Hill
2025-04-25 23:41:05 -07:00
committed by GitHub
parent 513f074766
commit df6f3ce883
21 changed files with 40 additions and 76 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
vllm/v1/worker/gpu_input_batch.py
View File

@@ -24,7 +24,6 @@ class CachedRequestState:
req_id: str
prompt_token_ids: list[int]
prompt: Optional[str]
mm_inputs: list[MultiModalKwargs]
mm_positions: list[PlaceholderRange]
sampling_params: SamplingParams

1
vllm/v1/worker/gpu_model_runner.py
View File

@@ -347,7 +347,6 @@ class GPUModelRunner(LoRAModelRunnerMixin):
self.requests[req_id] = CachedRequestState(
req_id=req_id,
prompt_token_ids=new_req_data.prompt_token_ids,
prompt=new_req_data.prompt,
mm_inputs=new_req_data.mm_inputs,
mm_positions=new_req_data.mm_positions,
sampling_params=sampling_params,

1
vllm/v1/worker/tpu_model_runner.py
View File

@@ -356,7 +356,6 @@ class TPUModelRunner:
self.requests[req_id] = CachedRequestState(
req_id=req_id,
prompt_token_ids=new_req_data.prompt_token_ids,
prompt=new_req_data.prompt,
mm_inputs=new_req_data.mm_inputs,
mm_positions=new_req_data.mm_positions,
sampling_params=sampling_params,