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

[v1] Pass BlockTable and KVCacheSpec to AttentionMetadataBuilders (#17483)

Browse Source 
Signed-off-by: Chen Zhang <zhangch99@outlook.com>
This commit is contained in:
Chen Zhang
2025-05-11 07:12:04 +08:00
committed by GitHub
parent 4c31218f80
commit 950751a987
11 changed files with 132 additions and 68 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
tests/v1/worker/test_gpu_input_batch.py
View File

@@ -221,6 +221,7 @@ def test_sampling_metadata_in_input_batch(device: str, batch_size: int):
max_num_reqs=batch_size,
max_model_len=1024,
max_num_blocks_per_req=10,
max_num_batched_tokens=1024,
device=torch.device(device),
pin_memory=is_pin_memory_available(),
vocab_size=1024,
@@ -310,6 +311,7 @@ def test_swap_states_in_input_batch(device: str, batch_size: int,
max_num_reqs=batch_size,
max_model_len=1024,
max_num_blocks_per_req=10,
max_num_batched_tokens=1024,
device=torch.device(device),
pin_memory=is_pin_memory_available(),
vocab_size=1024,
@@ -318,6 +320,7 @@ def test_swap_states_in_input_batch(device: str, batch_size: int,
max_num_reqs=batch_size,
max_model_len=1024,
max_num_blocks_per_req=10,
max_num_batched_tokens=1024,
device=torch.device(device),
pin_memory=is_pin_memory_available(),
vocab_size=1024,