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

[Core][Multimodal] Track encode cache entries by mm_hash and enable embedding sharing between requests (#22711)

Browse Source 
Signed-off-by: knlnguyen1802 <knlnguyen1802@gmail.com>
Signed-off-by: Roger Wang <hey@rogerw.io>
Co-authored-by: knlnguyen1802 <knlnguyen1802@gmail.com>
Co-authored-by: Roger Wang <hey@rogerw.io>
This commit is contained in:
Chenguang Zheng
2025-08-25 15:41:17 +08:00
committed by GitHub
parent 712d0f88d8
commit d765cf01fe
12 changed files with 365 additions and 154 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
tests/v1/worker/test_gpu_model_runner.py
View File

@@ -141,7 +141,7 @@ def _schedule_new_request(*req_ids: str) -> SchedulerOutput:
scheduled_encoder_inputs={},
num_common_prefix_blocks=0,
finished_req_ids=set(),
free_encoder_input_ids=[],
free_encoder_mm_hashes=[],
structured_output_request_ids={},
grammar_bitmask=None,
)
@@ -207,7 +207,7 @@ def test_update_states_request_finished(model_runner, dist_init):
scheduled_encoder_inputs={},
num_common_prefix_blocks=0,
finished_req_ids={req_id},
free_encoder_input_ids=[],
free_encoder_mm_hashes=[],
structured_output_request_ids={},
grammar_bitmask=None,
)
@@ -239,7 +239,7 @@ def test_update_states_request_resumed(model_runner, dist_init):
scheduled_encoder_inputs={},
num_common_prefix_blocks=0,
finished_req_ids=set(),
free_encoder_input_ids=[],
free_encoder_mm_hashes=[],
structured_output_request_ids={},
grammar_bitmask=None,
)
@@ -266,7 +266,7 @@ def test_update_states_request_resumed(model_runner, dist_init):
scheduled_encoder_inputs={},
num_common_prefix_blocks=0,
finished_req_ids=set(),
free_encoder_input_ids=[],
free_encoder_mm_hashes=[],
structured_output_request_ids={},
grammar_bitmask=None,
)
@@ -347,7 +347,7 @@ def test_update_states_no_changes(model_runner, dist_init):
scheduled_encoder_inputs={},
num_common_prefix_blocks=0,
finished_req_ids=set(),
free_encoder_input_ids=[],
free_encoder_mm_hashes=[],
structured_output_request_ids={},
grammar_bitmask=None,
)
@@ -384,7 +384,7 @@ def test_update_states_request_unscheduled(model_runner, dist_init):
scheduled_encoder_inputs={},
num_common_prefix_blocks=0,
finished_req_ids=set(),
free_encoder_input_ids=[],
free_encoder_mm_hashes=[],
structured_output_request_ids={},
grammar_bitmask=None,
)