ViperEkura c1d05ae11d perf: benchmark decode via real inference engine ...

- route decode benchmark through InferenceEngine generate path
- add enable_cuda_graph toggle to engine, scheduler, and executor
- make benchmark --cuda-graph/--no-cuda-graph control the toggle
- hoist local time imports to module top

2026-08-09 11:47:14 +08:00

..

cache

fix: isolate continuous batch decode state

2026-08-09 01:01:41 +08:00

network

refactor: split infer core into subpackages by concern

2026-08-08 23:43:05 +08:00

runtime

perf: benchmark decode via real inference engine

2026-08-09 11:47:14 +08:00

__init__.py

refactor: split infer core into subpackages by concern

2026-08-08 23:43:05 +08:00

engine.py

perf: benchmark decode via real inference engine

2026-08-09 11:47:14 +08:00

metrics.py

refactor: split infer core into subpackages by concern

2026-08-08 23:43:05 +08:00

scheduler.py

perf: benchmark decode via real inference engine

2026-08-09 11:47:14 +08:00

task.py

refactor: split infer core into subpackages by concern

2026-08-08 23:43:05 +08:00

workspace.py

refactor: split infer core into subpackages by concern

2026-08-08 23:43:05 +08:00