Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/LMCache/LMCache
/ functions
Functions
13,371 in github.com/LMCache/LMCache
⨍
Functions
13,371
◇
Types & classes
2,044
↳
Endpoints
181
↓ 1 callers
Method
make_key_bins
(self, config: CacheGenConfig)
lmcache/storage_backend/serde/cachegen_encoder.py:348
↓ 1 callers
Function
make_object_key
(chunk_hash: int, model_name: str = "test_model", kv_rank: int = 0)
tests/v1/distributed/test_report_status.py:109
↓ 1 callers
Function
make_storage_manager
(l1_size_mb: int = 256)
tests/v1/distributed/test_multimodel.py:58
↓ 1 callers
Method
make_value_bins
(self, config: CacheGenConfig)
lmcache/v1/storage_backend/naive_serde/cachegen_decoder.py:49
↓ 1 callers
Method
make_value_bins
(self, config: CacheGenConfig)
lmcache/v1/storage_backend/naive_serde/cachegen_encoder.py:34
↓ 1 callers
Method
make_value_bins
(self, config: CacheGenConfig)
lmcache/storage_backend/serde/cachegen_decoder.py:137
↓ 1 callers
Method
make_value_bins
(self, config: CacheGenConfig)
lmcache/storage_backend/serde/cachegen_encoder.py:354
↓ 1 callers
Method
master_for
( self, service_name, socket_timeout=None, username=None, password=None, **kwargs )
tests/conftest.py:253
↓ 1 callers
Method
matches
Return whether this rule applies to the layer's registered cache. Args: spec: The layer's vLLM KV cache spec (from its group).
lmcache/integration/vllm/kv_cache_group_edits.py:177
↓ 1 callers
Method
maybe_create_health_monitor
( self, lmcache_manager: "LMCacheManager" )
lmcache/v1/standalone/standalone_service_factory.py:83
↓ 1 callers
Method
maybe_create_internal_api_server
( self, lmcache_manager: "LMCacheManager" )
lmcache/v1/standalone/standalone_service_factory.py:78
↓ 1 callers
Method
maybe_create_lookup_client
(self)
lmcache/v1/standalone/standalone_service_factory.py:63
↓ 1 callers
Method
maybe_create_lookup_server
(self)
lmcache/v1/standalone/standalone_service_factory.py:69
↓ 1 callers
Method
maybe_create_offload_server
(self)
lmcache/v1/standalone/standalone_service_factory.py:72
↓ 1 callers
Method
maybe_create_runtime_plugin_launcher
(self)
lmcache/v1/standalone/standalone_service_factory.py:75
↓ 1 callers
Function
maybe_initialize_trace_recorder
Construct and register a trace recorder if configured. Args: bus: The active EventBus to subscribe the recorder to. obs_config: O
lmcache/v1/mp_observability/trace/lifecycle.py:40
↓ 1 callers
Method
mem_to_storage
( self, keys: Sequence[CacheEngineKey], mem_objs: List[MemoryObj] )
lmcache/v1/storage_backend/nixl_storage_backend.py:1089
↓ 1 callers
Method
memcheck
For debug purposes. Returns True is everything is fine, otherwise False.
lmcache/v1/memory_management.py:1813
↓ 1 callers
Method
metadata_container_offsets
Return checkpoint metadata container offsets in bytes.
lmcache/v1/storage_backend/plugins/rust_raw_block_backend.py:229
↓ 1 callers
Method
monitor_system
Monitor system metrics during benchmark
lmcache/tools/controller_benchmark/benchmark.py:586
↓ 1 callers
Method
move
Perform cross-node move of the KV cache.
lmcache/v1/cache_engine.py:1247
↓ 1 callers
Function
multi_layer_kv_transfer
Fully vectorized Python fallback for multi_layer_kv_transfer. Eliminates ALL token- and KV-level Python loops.
lmcache/python_ops_fallback.py:545
↓ 1 callers
Method
name
(self)
lmcache/cli/commands/bench/engine_bench/__init__.py:18
↓ 1 callers
Method
name
(self)
lmcache/cli/commands/tool/transfer_channel_benchmark.py:24
↓ 1 callers
Method
name
(self)
lmcache/cli/commands/tool/cache_simulator/sweep_command.py:16
↓ 1 callers
Method
name
(self)
lmcache/cli/commands/tool/cache_simulator/simulate_command.py:17
↓ 1 callers
Method
name
(self)
lmcache/cli/commands/tool/cache_simulator/gen_dataset_command.py:17
↓ 1 callers
Function
newTestScheme
newTestScheme returns a scheme with clientgo + the lmcache v1alpha1 types.
operator/internal/webhook/pod_injector_test.go:48
↓ 1 callers
Method
new_user_config
(user_id: int, workload_config: WorkloadConfig)
benchmarks/multi_round_qa/multi-round-qa.py:76
↓ 1 callers
Method
nixl_desc_exists
Check whether a NIXL descriptor exists in storage. :param meta_info: Descriptor key (file basename for FILE backends, ob
lmcache/v1/storage_backend/nixl_storage_backend.py:687
↓ 1 callers
Function
normalize_question
(question)
benchmarks/rag/utils.py:147
↓ 1 callers
Function
normalize_storage_manager_config
Normalize storage manager configuration. This consumes a matching DAX adapter as hybrid L1 Device-DAX overflow capacity. Args: c
lmcache/v1/distributed/config.py:257
↓ 1 callers
Function
notify_request
Notify a pending request that the KV store is complete. Returns True if the request was found and notified.
examples/disagg_prefill_mp/disagg_proxy_server.py:229
↓ 1 callers
Method
num_blocks
Return the number of paged blocks (raises for NBBS-fused formats).
lmcache/v1/gpu_connector/kv_format/specs/base.py:119
↓ 1 callers
Method
num_heads
Return the number of KV heads for ``layer_idx`` (1 for MLA).
lmcache/v1/gpu_connector/kv_format/specs/base.py:134
↓ 1 callers
Method
num_layers
Return the number of transformer layers.
lmcache/v1/gpu_connector/kv_format/specs/base.py:115
↓ 1 callers
Method
offload
( self, hashes: List[int], slot_mapping: List[int], offsets: List[int], )
lmcache/v1/offload_server/zmq_server.py:81
↓ 1 callers
Method
on_batched_get
(self, mock_objs: List[Optional[MockMemoryObj]])
lmcache/v1/storage_backend/connector/mock_connector.py:146
↓ 1 callers
Method
on_chunk_reuse
time_interval: float or int, in seconds
lmcache/observability.py:511
↓ 1 callers
Method
on_get
(self, mock_obj: MockMemoryObj)
lmcache/v1/storage_backend/connector/mock_connector.py:138
↓ 1 callers
Method
on_keys_created
(self, keys: list[ObjectKey])
lmcache/v1/distributed/eviction_policy/noop.py:35
↓ 1 callers
Method
on_keys_removed
(self, keys: list[ObjectKey])
lmcache/v1/distributed/eviction_policy/isolated_lru.py:92
↓ 1 callers
Method
on_keys_touched
(self, keys: list[ObjectKey])
lmcache/v1/distributed/eviction_policy/isolated_lru.py:83
↓ 1 callers
Method
on_l1_keys_finish_write_and_reserve_read
(self, keys: list[ObjectKey])
lmcache/v1/distributed/eviction.py:163
↓ 1 callers
Method
on_l1_keys_finish_write_and_reserve_read
(self, keys: list[ObjectKey])
lmcache/v1/distributed/storage_controllers/store_controller.py:147
↓ 1 callers
Method
on_l1_keys_reserved_read
(self, keys: list[ObjectKey])
lmcache/v1/distributed/eviction.py:142
↓ 1 callers
Method
on_l1_keys_reserved_write
(self, keys: list[ObjectKey])
lmcache/v1/distributed/eviction.py:149
↓ 1 callers
Method
on_l1_keys_write_finished
(self, keys: list[ObjectKey])
lmcache/v1/distributed/eviction.py:153
↓ 1 callers
Method
on_l2_keys_accessed
(self, keys: list[ObjectKey])
lmcache/v1/distributed/eviction.py:188
↓ 1 callers
Method
on_l2_keys_deleted
(self, keys: list[ObjectKey])
lmcache/v1/distributed/eviction.py:191
↓ 1 callers
Method
on_p2p_transfer_finished
(self, request_id: int)
lmcache/observability.py:503
↓ 1 callers
Method
on_p2p_transfer_request
(self, num_tokens: int)
lmcache/observability.py:491
↓ 1 callers
Method
on_put
(self, mock_obj: MockMemoryObj)
lmcache/v1/storage_backend/connector/mock_connector.py:130
↓ 1 callers
Method
on_request_finished
Called when a request finishes (from the loop thread). Stateless workloads can implement this as a no-op. Stateful workloads use it t
lmcache/cli/commands/bench/engine_bench/workloads/base.py:63
↓ 1 callers
Method
on_request_store_finished
Callback when request finishes AND all its KV cache store ops completes. This method ensures that request_ids_set is not empty.
lmcache/integration/request_telemetry/base.py:29
↓ 1 callers
Method
on_server_message
(self, message)
examples/blend_kv/online_kv.py:95
↓ 1 callers
Method
on_user_message
(self, message)
examples/blend_kv/online_kv.py:92
↓ 1 callers
Function
ordinal_word
0 -> 'first', 1 -> 'second', ...
benchmarks/multi_doc_qa/shuffle_doc_qa.py:32
↓ 1 callers
Function
override_config_from_dict
Override configuration with dictionary
lmcache/v1/standalone/__main__.py:376
↓ 1 callers
Method
page_buffer_size
Return ``num_blocks * block_size`` (or the fused PBS axis). Non-MP only (see the class docstring).
lmcache/v1/gpu_connector/kv_format/specs/base.py:127
↓ 1 callers
Function
parse_args
()
tests/v1/storage_backend/test_eic.py:22
↓ 1 callers
Function
parse_args
()
tests/v1/shm_allocator/verify_shmfile_tcp.py:39
↓ 1 callers
Function
parse_args
()
examples/disagg_prefill_mp/disagg_proxy_server.py:110
↓ 1 callers
Function
parse_args
()
examples/online_session/openai_chat_completion_client.py:146
↓ 1 callers
Function
parse_args
Parse command-line arguments for the script. Returns: argparse.Namespace: Parsed arguments containing: - disk_path: Path to t
examples/kv_cache_reuse/local_backends/rust_backend_offload.py:119
↓ 1 callers
Function
parse_args
()
examples/kv_cache_reuse/local_backends/offload.py:96
↓ 1 callers
Function
parse_args
()
examples/disagg_prefill/disagg_proxy_server.py:274
↓ 1 callers
Function
parse_args
()
examples/blend_kv_v1/blend.py:103
↓ 1 callers
Function
parse_args
Parse command-line arguments
lmcache/v1/standalone/__main__.py:418
↓ 1 callers
Function
parse_args
()
lmcache/v1/multiprocess/http_server.py:257
↓ 1 callers
Function
parse_args
Parse command line arguments for the cache server. Returns: Parsed arguments namespace.
lmcache/v1/multiprocess/server.py:415
↓ 1 callers
Function
parse_args
Parse command-line arguments.
benchmarks/musa/bench_inprocess_transfer.py:55
↓ 1 callers
Function
parse_args
()
benchmarks/ttft-estimator/ttft-estimator.py:90
↓ 1 callers
Function
parse_args_to_p2p_config
Convert parsed command line arguments to a P2PConfig. Args: args: Parsed arguments from the argument parser. Returns: The co
lmcache/v1/multiprocess/config.py:460
↓ 1 callers
Function
parse_arguments
Parse command-line arguments.
examples/agents/prefix_analysis.py:359
↓ 1 callers
Function
parse_arguments
()
benchmarks/multi_round_qa/multi-round-qa.py:578
↓ 1 callers
Function
parse_arguments
()
benchmarks/rag/precompute.py:133
↓ 1 callers
Function
parse_arguments
()
benchmarks/rag/rag.py:82
↓ 1 callers
Function
parse_chunk_output
(choice)
benchmarks/multi_doc_qa/shuffle_doc_qa.py:146
↓ 1 callers
Function
parse_generation
(s)
benchmarks/rag/utils.py:154
↓ 1 callers
Function
parse_kv_shape
Parse KV shape from string like '32,2,256,32,128
lmcache/v1/standalone/__main__.py:393
↓ 1 callers
Function
parse_kvcache_shape_spec
Parse KV shape specification with multiple layer groups. Format examples: - "(2,2,256,4,16):float16:2" (single group) - "(2,2,256,4,16):f
lmcache/v1/standalone/__main__.py:81
↓ 1 callers
Function
parse_mp_config
Parse the aggregated MP config from the environment. Returns: A dict with keys like ``mp_config``, ``storage_manager_config``, ``
examples/mp_runtime_plugins/mp_plugin.py:48
↓ 1 callers
Function
parse_pool_sizes
( pool_sizes_input: Optional[List[str]], )
examples/agents/prefix_analysis.py:415
↓ 1 callers
Function
parse_process_summary
()
benchmarks/multi_round_qa/multi-round-qa.py:678
↓ 1 callers
Function
parse_prompt_build_method
( prompt_build_method: str, )
benchmarks/rag/precompute.py:200
↓ 1 callers
Function
parse_size
(size: str)
benchmarks/rag/precompute.py:181
↓ 1 callers
Function
pickEphemeralPort
pickEphemeralPort asks the kernel for a free TCP port on 127.0.0.1 and immediately releases it. There is a small race window between release and kubec
operator/test/utils/portforward.go:132
↓ 1 callers
Function
pick_derangements
Return `count` distinct derangements; raises if mathematically impossible.
benchmarks/multi_doc_qa/shuffle_doc_qa.py:70
↓ 1 callers
Function
pick_up_bound_clients
(client_id: str)
examples/disagg_prefill/disagg_proxy_server.py:452
↓ 1 callers
Method
pin
Execute a pin cache operation with error handling.
lmcache/v1/cache_controller/executor.py:109
↓ 1 callers
Method
pin
( self, key: CacheEngineKey, )
lmcache/v1/storage_backend/p2p_backend.py:773
↓ 1 callers
Method
pin
(self, key: CacheEngineKey)
lmcache/v1/storage_backend/plugins/rust_raw_block_backend.py:358
↓ 1 callers
Method
ping
(self)
tests/v1/test_health_monitor_fallback_recovery.py:48
↓ 1 callers
Function
plot_hit_rates
Generate and save the hit rate vs pool size plot comparing both methods.
examples/agents/prefix_analysis.py:282
↓ 1 callers
Function
plot_statistics
Render and save a 2×4 multi-panel figure with seven statistical charts. Parameters ---------- results: Output of :func:`simu
lmcache/tools/cache_simulator/simulator.py:453
↓ 1 callers
Method
pop_completed_store_tasks
(self)
lmcache/v1/distributed/l2_adapters/nixl_store_dynamic_l2_adapter.py:414
↓ 1 callers
Method
pop_completed_store_tasks
Pop all completed store tasks. Returns: dict[L2TaskId, L2StoreResult]: a dictionary mapping the task id to an ``L2Sto
lmcache/v1/distributed/l2_adapters/nixl_store_l2_adapter.py:515
↓ 1 callers
Method
pop_completed_store_tasks
( self, )
lmcache/v1/distributed/l2_adapters/native_connector_l2_adapter.py:214
↓ 1 callers
Method
pop_completed_store_tasks
(self)
lmcache/v1/distributed/l2_adapters/serde_wrapper.py:216
← previous
next →
4,401–4,500 of 13,371, ranked by callers