Huazhong Ji 8ea8523744 reset default block_size from 16 to 128 (#84) ...

### What this PR does / why we need it?

Changed default block_size in platform.py from 16 to 128, as Ascend
Devices have a better affinity for block size 128.

### Does this PR introduce _any_ user-facing change?
No

### How was this patch tested?
CI passed

Signed-off-by: hzji210@gmail.com <hzji210@gmail.com>

2025-02-18 14:19:38 +08:00

..

ops

[Core] Init vllm-ascend (#3)

2025-02-05 10:53:12 +08:00

__init__.py

[Core] Init vllm-ascend (#3)

2025-02-05 10:53:12 +08:00

attention.py

[attn] fix device of tensors in attention (#25)

2025-02-10 19:20:29 +08:00

communicator.py

[dist] revert communicator patch (#66)

2025-02-17 11:42:33 +08:00

model_runner.py

bugfix for mrope (#14)

2025-02-07 16:46:24 +08:00

platform.py

reset default block_size from 16 to 128 (#84)

2025-02-18 14:19:38 +08:00

utils.py

[Worker] Register mindie_turbo while initializing NPUWorker (#13)

2025-02-07 16:47:17 +08:00

worker.py

[dist] revert communicator patch (#66)

2025-02-17 11:42:33 +08:00