Mooncake/mooncake-integration
zbtrs2 ddcc56c2f6 [Doc] ollama integration: showcase README, report figures, demo scoreboard
Lead the README with the scaling result and a '6 agents, one prefill' hero,
add a 'what's new' contribution table, and embed the architecture and result
figures into the README and report. demo.sh now prints a one-glance before/after
scoreboard (raw stats behind DEMO_VERBOSE), captured in docs/demo_scoreboard.txt.

Metric framing is kept unambiguous: 'total prefill work avoided' and 'redundant
prefill eliminated' are reported as distinct numbers, and warm-store vs
cold-start regimes are labelled so they cannot be misread as conflicting.

Co-Authored-By: Claude <noreply@anthropic.com>
2026-06-24 23:10:45 +08:00
..
ollama [Doc] ollama integration: showcase README, report figures, demo scoreboard 2026-06-24 23:10:45 +08:00
store [Store] Expose batch_replica_clear in Python binding (#1848) 2026-04-17 17:36:32 +08:00
transfer_engine [TransferEngine][ROCm] Add ROCm HIP support to the Mooncake Python package (#1742) 2026-04-13 11:13:10 +08:00
CMakeLists.txt [Docker] fix(docker): respect PYTHON_VERSION build-arg when building wheel (#1745) 2026-04-10 09:56:19 +08:00
allocator.py [TE] Add early mem backend detection method in NVLINK_allocator (#1296) 2026-01-04 15:20:52 +08:00
allocator_ascend_npu.py [TE] Ubshmem transport support ipc memory and build allocator when set USE_UBSHMEM=ON (#1519) 2026-02-14 19:30:06 +08:00
integration_utils.h [Store] validate metadata when put_tensor (#1396) 2026-01-26 16:31:48 +08:00