CMakeLists.txt
LICENSE
MANIFEST.in
README.md
pyproject.toml
requirements.txt
setup.py
benchmarks/.gitkeep
benchmarks/README.md
benchmarks/trace_replay.py
examples/offline_inference.py
examples/offline_inference_esa.py
examples/offline_inference_kvcomp.py
examples/offline_inference_kvstar.py
examples/ucm_config_example.yaml
examples/metrics/grafana.json
examples/metrics/metrics_configs.yaml
test/test_mooncake.py
test/test_ucm_connector_save_load.py
uc_manager.egg-info/PKG-INFO
uc_manager.egg-info/SOURCES.txt
uc_manager.egg-info/dependency_links.txt
uc_manager.egg-info/not-zip-safe
uc_manager.egg-info/top_level.txt
ucm/CMakeLists.txt
ucm/__init__.py
ucm/logger.py
ucm/utils.py
ucm/integration/__init__.py
ucm/integration/vllm/__init__.py
ucm/integration/vllm/ucm_connector.py
ucm/integration/vllm/patch/__init__.py
ucm/integration/vllm/patch/apply_patch.py
ucm/integration/vllm/patch/0.9.1/vllm-adapt.patch
ucm/integration/vllm/patch/0.9.1/vllm-ascend-adapt.patch
ucm/integration/vllm/patch/0.9.2/vllm-adapt-aggre.patch
ucm/integration/vllm/patch/0.9.2/vllm-adapt-pc.patch
ucm/integration/vllm/patch/0.9.2/vllm-adapt-sparse.patch
ucm/integration/vllm/patch/0.9.2/vllm-ascend-adapt.patch
ucm/integration/vllm/patch/patch_funcs/__init__.py
ucm/integration/vllm/patch/patch_funcs/v092/__init__.py
ucm/integration/vllm/patch/patch_funcs/v092/vllm_ascend_patch.py
ucm/integration/vllm/patch/patch_funcs/v092/vllm_patch.py
ucm/pd/__init__.py
ucm/pd/toy_proxy_server.py
ucm/sandbox/agentic_ai/README.md
ucm/sandbox/sparse/retake/.gitignore
ucm/sandbox/sparse/retake/.gitkeep
ucm/sandbox/sparse/retake/README.md
ucm/sandbox/sparse/retake/demo.py
ucm/sandbox/sparse/retake/environment_npu.yaml
ucm/sandbox/sparse/retake/requirements.txt
ucm/sandbox/sparse/retake/configs/demo.yaml
ucm/sandbox/sparse/retake/configs/demo_npu.yaml
ucm/sandbox/sparse/retake/configs/llava_video/llava-video_lvbench.yaml
ucm/sandbox/sparse/retake/configs/llava_video/llava-video_mlvu.yaml
ucm/sandbox/sparse/retake/configs/llava_video/llava-video_videomme.yaml
ucm/sandbox/sparse/retake/configs/llava_video/retake_llava-video_lvbench.yaml
ucm/sandbox/sparse/retake/configs/llava_video/retake_llava-video_mlvu.yaml
ucm/sandbox/sparse/retake/configs/llava_video/retake_llava-video_videomme.yaml
ucm/sandbox/sparse/retake/configs/qwen2_5_vl/flexreduc_qwen2-5-vl_longvideobench.yaml
ucm/sandbox/sparse/retake/configs/qwen2_5_vl/flexreduc_qwen2-5-vl_lvbench.yaml
ucm/sandbox/sparse/retake/configs/qwen2_5_vl/flexreduc_qwen2-5-vl_mlvu.yaml
ucm/sandbox/sparse/retake/configs/qwen2_5_vl/flexreduc_qwen2-5-vl_videomme.yaml
ucm/sandbox/sparse/retake/configs/qwen2_5_vl/qwen2-5-vl_videomme_f256.yaml
ucm/sandbox/sparse/retake/configs/qwen2_vl/qwen2-vl_lvbench.yaml
ucm/sandbox/sparse/retake/configs/qwen2_vl/qwen2-vl_mlvu.yaml
ucm/sandbox/sparse/retake/configs/qwen2_vl/qwen2-vl_videomme.yaml
ucm/sandbox/sparse/retake/configs/qwen2_vl/retake_qwen2-vl_lvbench.yaml
ucm/sandbox/sparse/retake/configs/qwen2_vl/retake_qwen2-vl_mlvu.yaml
ucm/sandbox/sparse/retake/configs/qwen2_vl/retake_qwen2-vl_videomme.yaml
ucm/sandbox/sparse/retake/docs/prepare_longvideobench.md
ucm/sandbox/sparse/retake/docs/prepare_lvbench.md
ucm/sandbox/sparse/retake/docs/prepare_mlvu.md
ucm/sandbox/sparse/retake/docs/prepare_videomme.md
ucm/sandbox/sparse/retake/misc/flexreduc_pipeline.png
ucm/sandbox/sparse/retake/retake/dataset_utils.py
ucm/sandbox/sparse/retake/retake/infer_eval.py
ucm/sandbox/sparse/retake/retake/llava_onevision.py
ucm/sandbox/sparse/retake/retake/longvideo_cache.py
ucm/sandbox/sparse/retake/retake/monkeypatch.py
ucm/sandbox/sparse/retake/retake/qwen2_5_vl.py
ucm/sandbox/sparse/retake/retake/qwen2_vl.py
ucm/sandbox/sparse/retake/retake/visual_compression.py
ucm/sandbox/sparse/retake/scripts/infer_eval.sh
ucm/sandbox/sparse/retake/scripts/submission/prepare_lvbench_submission.py
ucm/sandbox/sparse/retake/scripts/submission/prepare_mlvu_test_submission.py
ucm/sandbox/sparse/retake/scripts/submission/prepare_videomme_submission.py
ucm/sandbox/sparse/retake/scripts/utils/build_longvideobench_dataset.py
ucm/sandbox/sparse/retake/scripts/utils/build_lvbench_dataset.py
ucm/sandbox/sparse/retake/scripts/utils/build_mlvu_dataset.py
ucm/sandbox/sparse/retake/scripts/utils/build_mlvu_test_dataset.py
ucm/sandbox/sparse/retake/scripts/utils/build_videomme_dataset.py
ucm/sandbox/sparse/retake/scripts/utils/convert_llava_video_weights_to_hf.py
ucm/sandbox/sparse/retake/scripts/utils/frame_extraction.py
ucm/shared/CMakeLists.txt
ucm/shared/__init__.py
ucm/shared/infra/CMakeLists.txt
ucm/shared/infra/logger/logger.h
ucm/shared/infra/logger/flux/flux_logger.cc
ucm/shared/infra/logger/spdlog/spdlog_logger.cc
ucm/shared/infra/status/status.h
ucm/shared/infra/template/hashset.h
ucm/shared/infra/template/singleton.h
ucm/shared/infra/template/timer.h
ucm/shared/infra/template/topn_heap.h
ucm/shared/infra/thread/index_pool.h
ucm/shared/infra/thread/latch.h
ucm/shared/infra/thread/thread_pool.h
ucm/shared/infra/time/stopwatch.h
ucm/shared/metrics/CMakeLists.txt
ucm/shared/metrics/__init__.py
ucm/shared/metrics/observability.py
ucm/shared/metrics/cc/stats_monitor.cc
ucm/shared/metrics/cc/stats_monitor.h
ucm/shared/metrics/cc/stats_registry.cc
ucm/shared/metrics/cc/stats_registry.h
ucm/shared/metrics/cc/stats/conn_stats.cc
ucm/shared/metrics/cc/stats/conn_stats.h
ucm/shared/metrics/cc/stats/istats.h
ucm/shared/metrics/cpy/metrics.py.cc
ucm/shared/metrics/test/test.py
ucm/shared/test/CMakeLists.txt
ucm/shared/test/case/infra/hashset_test.cc
ucm/shared/test/case/trans/trans_test.cc
ucm/shared/test/example/trans/trans_on_cuda_example.py
ucm/shared/trans/CMakeLists.txt
ucm/shared/trans/__init__.py
ucm/shared/trans/buffer.h
ucm/shared/trans/device.h
ucm/shared/trans/stream.h
ucm/shared/trans/ascend/CMakeLists.txt
ucm/shared/trans/ascend/ascend_buffer.cc
ucm/shared/trans/ascend/ascend_buffer.h
ucm/shared/trans/ascend/ascend_device.cc
ucm/shared/trans/ascend/ascend_stream.cc
ucm/shared/trans/ascend/ascend_stream.h
ucm/shared/trans/cpy/trans.py.cc
ucm/shared/trans/cuda/CMakeLists.txt
ucm/shared/trans/cuda/cuda_buffer.cc
ucm/shared/trans/cuda/cuda_buffer.h
ucm/shared/trans/cuda/cuda_device.cc
ucm/shared/trans/cuda/cuda_sm_kernel.cu
ucm/shared/trans/cuda/cuda_sm_kernel.h
ucm/shared/trans/cuda/cuda_sm_stream.cc
ucm/shared/trans/cuda/cuda_sm_stream.h
ucm/shared/trans/cuda/cuda_stream.cc
ucm/shared/trans/cuda/cuda_stream.h
ucm/shared/trans/detail/indexer.h
ucm/shared/trans/detail/reserved_buffer.h
ucm/shared/trans/maca/CMakeLists.txt
ucm/shared/trans/maca/maca_sm_kernel.cu
ucm/shared/trans/simu/CMakeLists.txt
ucm/shared/trans/simu/simu_buffer.cc
ucm/shared/trans/simu/simu_buffer.h
ucm/shared/trans/simu/simu_device.cc
ucm/shared/trans/simu/simu_stream.cc
ucm/shared/trans/simu/simu_stream.h
ucm/shared/vendor/CMakeLists.txt
ucm/sparse/CMakeLists.txt
ucm/sparse/__init__.py
ucm/sparse/base.py
ucm/sparse/factory.py
ucm/sparse/state.py
ucm/sparse/utils.py
ucm/sparse/cache_blend/README.md
ucm/sparse/esa/CMakeLists.txt
ucm/sparse/esa/__init__.py
ucm/sparse/esa/esa.py
ucm/sparse/esa/retrieval/CMakeLists.txt
ucm/sparse/esa/retrieval/__init__.py
ucm/sparse/esa/retrieval/retrieval_worker.py
ucm/sparse/esa/retrieval/cpy/retrieval_backend.cpp
ucm/sparse/gsa/CMakeLists.txt
ucm/sparse/gsa/__init__.py
ucm/sparse/gsa/gsa.py
ucm/sparse/gsa/offload_ops/CMakeLists.txt
ucm/sparse/gsa/offload_ops/__init__.py
ucm/sparse/gsa/offload_ops/include/cal_kpre_and_topk.h
ucm/sparse/gsa/offload_ops/include/k_repre.h
ucm/sparse/gsa/offload_ops/include/select_topk_block.h
ucm/sparse/gsa/offload_ops/include/thread_safe_queue.h
ucm/sparse/gsa/offload_ops/include/vec_product.h
ucm/sparse/gsa/offload_ops/src/cal_kpre_and_topk.cpp
ucm/sparse/gsa/offload_ops/src/k_repre.cpp
ucm/sparse/gsa/offload_ops/src/pybinds.cpp
ucm/sparse/gsa/offload_ops/src/select_topk_block.cpp
ucm/sparse/gsa/offload_ops/src/thread_safe_queue.cpp
ucm/sparse/gsa/offload_ops/src/vec_product.cpp
ucm/sparse/gsa/prefetch/CMakeLists.txt
ucm/sparse/gsa/prefetch/__init__.py
ucm/sparse/gsa/prefetch/prefetch_engine.py
ucm/sparse/gsa/prefetch/include/kvcache_log.h
ucm/sparse/gsa/prefetch/include/kvcache_pre.h
ucm/sparse/gsa/prefetch/src/kvcache_pre.cpp
ucm/sparse/gsa/prefetch/src/pybinds.cpp
ucm/sparse/kvcomp/.gitkeep
ucm/sparse/kvcomp/CMakeLists.txt
ucm/sparse/kvcomp/README.md
ucm/sparse/kvcomp/__init__.py
ucm/sparse/kvcomp/hash_encoder.py
ucm/sparse/kvcomp/kvcomp.py
ucm/sparse/kvcomp/kvcomp_config.py
ucm/sparse/kvcomp/offline_inference_kvcomp.py
ucm/sparse/kvcomp/configs/kvcomp_deepseek_v2_lite_config.json
ucm/sparse/kvcomp/configs/kvcomp_qwen3_32B_config.json
ucm/sparse/kvcomp/configs/kvcomp_qwen3_4B_config.json
ucm/sparse/kvcomp/configs/kvcomp_qwq_32B_config.json
ucm/sparse/kvcomp/figs/kvcomp_end_to_end_performance.jpg
ucm/sparse/kvcomp/figs/kvcomp_longbench.jpg
ucm/sparse/kvcomp/figs/kvcomp_scheme.jpg
ucm/sparse/kvcomp/figs/kvcomp_single_layer_performance.jpg
ucm/sparse/kvcomp/hash_retrieval/CMakeLists.txt
ucm/sparse/kvcomp/hash_retrieval/__init__.py
ucm/sparse/kvcomp/hash_retrieval/hash_retrieval_worker.py
ucm/sparse/kvcomp/hash_retrieval/cpy/hash_retrieval_backend.cpp
ucm/sparse/kvcomp/paper/kvcomp-ACL-2025-paper.pdf
ucm/sparse/kvstar/.gitkeep
ucm/sparse/kvstar/CMakeLists.txt
ucm/sparse/kvstar/__init__.py
ucm/sparse/kvstar/multistep.py
ucm/sparse/kvstar/utils.py
ucm/sparse/kvstar/retrieve/CMakeLists.txt
ucm/sparse/kvstar/retrieve/__init__.py
ucm/sparse/kvstar/retrieve/core/CMakeLists.txt
ucm/sparse/kvstar/retrieve/core/api/kvstar_retrieve/kvstar_retrieve.cpp
ucm/sparse/kvstar/retrieve/core/api/kvstar_retrieve/kvstar_retrieve.h
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/computation_task.h
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/retrieve_task.h
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/retrieve_task_manager.cpp
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/retrieve_task_manager.h
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/retrieve_task_queue.cpp
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/retrieve_task_queue.h
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/retrieve_task_runner.cpp
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/retrieve_task_runner.h
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/retrieve_task_set.h
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/retrieve_task_waiter.h
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/simd_compute_kernel.cpp
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/simd_compute_kernel.h
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/task_result.h
ucm/sparse/kvstar/retrieve/core/domain/retrieve_task/task_status.h
ucm/sparse/kvstar/retrieve/core/infra/logger/logger.cpp
ucm/sparse/kvstar/retrieve/core/infra/logger/logger.h
ucm/sparse/kvstar/retrieve/core/infra/memory/memory.cpp
ucm/sparse/kvstar/retrieve/core/infra/memory/memory.h
ucm/sparse/kvstar/retrieve/core/infra/status/status.h
ucm/sparse/kvstar/retrieve/core/infra/template/singleton.h
ucm/sparse/kvstar/retrieve/core/infra/thread/latch.h
ucm/sparse/kvstar/retrieve/py_intf/CMakeLists.txt
ucm/sparse/kvstar/retrieve/py_intf/py_intf.cpp
ucm/store/CMakeLists.txt
ucm/store/__init__.py
ucm/store/factory.py
ucm/store/ucmstore.h
ucm/store/ucmstore.py
ucm/store/device/CMakeLists.txt
ucm/store/device/ibuffered_device.h
ucm/store/device/idevice.h
ucm/store/device/ascend/CMakeLists.txt
ucm/store/device/ascend/ascend_device.cc
ucm/store/device/cuda/CMakeLists.txt
ucm/store/device/cuda/cuda_device.cu
ucm/store/device/maca/CMakeLists.txt
ucm/store/device/maca/maca_device.cu
ucm/store/device/musa/CMakeLists.txt
ucm/store/device/musa/musa_device.cc
ucm/store/device/musa/musa_device.mu
ucm/store/device/simu/CMakeLists.txt
ucm/store/device/simu/simu_device.cc
ucm/store/dramstore/CMakeLists.txt
ucm/store/dramstore/__init__.py
ucm/store/dramstore/dramstore_connector.py
ucm/store/infra/CMakeLists.txt
ucm/store/infra/file/file.cc
ucm/store/infra/file/file.h
ucm/store/infra/file/ifile.h
ucm/store/infra/file/posix_file.cc
ucm/store/infra/file/posix_file.h
ucm/store/localstore/CMakeLists.txt
ucm/store/localstore/__init__.py
ucm/store/localstore/cc/api/localstore.cc
ucm/store/localstore/cc/api/localstore.h
ucm/store/localstore/cc/domain/cache/cache_data.cc
ucm/store/localstore/cc/domain/cache/cache_data.h
ucm/store/localstore/cc/domain/cache/cache_hash.cc
ucm/store/localstore/cc/domain/cache/cache_hash.h
ucm/store/localstore/cc/domain/cache/cache_index.cc
ucm/store/localstore/cc/domain/cache/cache_index.h
ucm/store/localstore/cc/domain/cache/cache_instance.cc
ucm/store/localstore/cc/domain/cache/cache_instance.h
ucm/store/localstore/cc/domain/cache/cache_layout.cc
ucm/store/localstore/cc/domain/cache/cache_layout.h
ucm/store/localstore/cc/domain/cache/cache_meta.cc
ucm/store/localstore/cc/domain/cache/cache_meta.h
ucm/store/localstore/cc/domain/cache/cache_segment.cc
ucm/store/localstore/cc/domain/cache/cache_segment.h
ucm/store/localstore/cpy/localstore.py.cc
ucm/store/mooncakestore/CMakeLists.txt
ucm/store/mooncakestore/__init__.py
ucm/store/mooncakestore/mooncake_connector.py
ucm/store/nfsstore/CMakeLists.txt
ucm/store/nfsstore/__init__.py
ucm/store/nfsstore/nfsstore_connector.py
ucm/store/nfsstore/cc/api/nfsstore.cc
ucm/store/nfsstore/cc/api/nfsstore.h
ucm/store/nfsstore/cc/domain/hotness/hotness_manager.h
ucm/store/nfsstore/cc/domain/hotness/hotness_set.cc
ucm/store/nfsstore/cc/domain/hotness/hotness_set.h
ucm/store/nfsstore/cc/domain/hotness/hotness_timer.h
ucm/store/nfsstore/cc/domain/space/space_layout.h
ucm/store/nfsstore/cc/domain/space/space_manager.cc
ucm/store/nfsstore/cc/domain/space/space_manager.h
ucm/store/nfsstore/cc/domain/space/space_property.cc
ucm/store/nfsstore/cc/domain/space/space_property.h
ucm/store/nfsstore/cc/domain/space/space_recycle.cc
ucm/store/nfsstore/cc/domain/space/space_recycle.h
ucm/store/nfsstore/cc/domain/space/space_shard_layout.cc
ucm/store/nfsstore/cc/domain/space/space_shard_layout.h
ucm/store/nfsstore/cc/domain/space/space_shard_temp_layout.cc
ucm/store/nfsstore/cc/domain/space/space_shard_temp_layout.h
ucm/store/nfsstore/cc/domain/trans/posix_queue.cc
ucm/store/nfsstore/cc/domain/trans/posix_queue.h
ucm/store/nfsstore/cc/domain/trans/trans_manager.h
ucm/store/nfsstore/cpy/nfsstore.py.cc
ucm/store/pcstore/CMakeLists.txt
ucm/store/pcstore/__init__.py
ucm/store/pcstore/pcstore_connector.py
ucm/store/pcstore/cc/api/pcstore.cc
ucm/store/pcstore/cc/api/pcstore.h
ucm/store/pcstore/cc/domain/space/space_layout.cc
ucm/store/pcstore/cc/domain/space/space_layout.h
ucm/store/pcstore/cc/domain/space/space_manager.cc
ucm/store/pcstore/cc/domain/space/space_manager.h
ucm/store/pcstore/cc/domain/trans/share_buffer.cc
ucm/store/pcstore/cc/domain/trans/share_buffer.h
ucm/store/pcstore/cc/domain/trans/trans_manager.cc
ucm/store/pcstore/cc/domain/trans/trans_manager.h
ucm/store/pcstore/cc/domain/trans/trans_queue.cc
ucm/store/pcstore/cc/domain/trans/trans_queue.h
ucm/store/pcstore/cc/domain/trans/trans_share_queue.cc
ucm/store/pcstore/cc/domain/trans/trans_share_queue.h
ucm/store/pcstore/cc/domain/trans/trans_task.h
ucm/store/pcstore/cpy/pcstore.py.cc
ucm/store/task/CMakeLists.txt
ucm/store/task/task_manager.cc
ucm/store/task/task_manager.h
ucm/store/task/task_queue.h
ucm/store/task/task_set.h
ucm/store/task/task_shard.h
ucm/store/task/task_waiter.h
ucm/store/test/CMakeLists.txt
ucm/store/test/case/cmn/path_base.h
ucm/store/test/case/cmn/random.h
ucm/store/test/case/infra/file_test.cc
ucm/store/test/case/infra/posix_file_test.cc
ucm/store/test/case/localstore/cache_hash_test.cc
ucm/store/test/case/localstore/cache_index_test.cc
ucm/store/test/case/localstore/cache_instance_test.cc
ucm/store/test/case/localstore/cache_layout_test.cc
ucm/store/test/case/localstore/cache_segment_test.cc
ucm/store/test/case/nfsstore/hotness_test.cc
ucm/store/test/case/nfsstore/space_manager_test.cc
ucm/store/test/case/nfsstore/space_property_test.cc
ucm/store/test/case/nfsstore/space_recycle_test.cc
ucm/store/test/e2e/dramstore_embed_and_fetch.py
ucm/store/test/e2e/nfsstore_embed.py
ucm/store/test/e2e/nfsstore_embed_fetch.py
ucm/store/test/e2e/nfsstore_embed_fetch_run.py
ucm/store/test/e2e/nfsstore_fetch.py
ucm/store/test/e2e/pcstore_embed.py
ucm/store/test/e2e/pcstore_fetch.py