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
↓ 2 callers
Method
from_extra_config
( cls, extra_config: Dict[str, Any], plugin_name: Optional[str] = None )
lmcache/v1/storage_backend/connector/bigtable_config.py:27
↓ 2 callers
Method
from_gpu
(self, memory_obj: MemoryObj, start: int, end: int, **kwargs)
lmcache/v1/gpu_connector/gpu_connectors.py:569
↓ 2 callers
Method
from_gpu
(self, memory_obj: MemoryObj, start: int, end: int, **kwargs)
lmcache/v1/gpu_connector/xpu_connectors.py:406
↓ 2 callers
Method
from_model_name
(model_name: str)
lmcache/v1/storage_backend/naive_serde/cachegen_basics.py:27
↓ 2 callers
Method
from_model_name
(model_name: str)
lmcache/storage_backend/serde/cachegen_basics.py:42
↓ 2 callers
Function
generate_kv_cache_paged
(num_blocks, device, block_size=16, dtype=torch.bfloat16)
tests/v1/utils.py:371
↓ 2 callers
Function
generate_kv_cache_paged_list_tensors
Instead of Tuple[Tuple[Tensor, Tensor]], return List[Tensor] where KV are in the same tensor
tests/disagg/test_nixl_cache_engine.py:31
↓ 2 callers
Function
generate_stream
()
examples/disagg_prefill/disagg_proxy_server.py:538
↓ 2 callers
Function
generate_test_data
( num_objs: int, shape: torch.Size, dtype: torch.dtype = torch.bfloat16 )
tests/disagg/test_nixl_storage_backend.py:32
↓ 2 callers
Method
get
Retrieve a memory object from Valkey by key.
lmcache/v1/storage_backend/connector/valkey_connector.py:405
↓ 2 callers
Method
get
Get the memory_obj of the corresponding key Input: key: the key of the corresponding object Returns:
lmcache/v1/storage_backend/connector/base_connector.py:175
↓ 2 callers
Method
get
(self, key: CacheEngineKey)
lmcache/v1/storage_backend/connector/s3_connector.py:341
↓ 2 callers
Method
get
Get data from file system
lmcache/v1/storage_backend/connector/fs_connector.py:224
↓ 2 callers
Function
getFirstFoundEnvTestBinaryDir
getFirstFoundEnvTestBinaryDir locates the first binary in the specified path. ENVTEST-based tests depend on specific binaries, usually located in path
operator/internal/controller/suite_test.go:109
↓ 2 callers
Function
get_abstract_methods
Get all abstract methods from a class.
tests/v1/utils.py:698
↓ 2 callers
Method
get_active_count
Get the count of active threads.
lmcache/v1/periodic_thread.py:494
↓ 2 callers
Function
get_all_methods_from_base
Get all public methods defined in the base class (excluding inherited from object).
tests/v1/utils.py:637
↓ 2 callers
Method
get_all_worker_infos
Get WorkerInfo for all workers in this instance.
lmcache/v1/cache_controller/utils.py:287
↓ 2 callers
Function
get_backend_cname
(backend: StorageBackendInterface)
lmcache/v1/storage_backend/storage_manager.py:59
↓ 2 callers
Method
get_block_ids_with_load_errors
(self)
lmcache/integration/vllm/vllm_v1_adapter.py:1334
↓ 2 callers
Method
get_by_level
Get all threads with the specified level.
lmcache/v1/periodic_thread.py:484
↓ 2 callers
Method
get_cached
Get cached data, refreshing if expired. Args: get_fresh_data: Function to get fresh data when cache is expired
lmcache/v1/utils/cache_utils.py:30
↓ 2 callers
Method
get_client
(self)
lmcache/v1/storage_backend/connector/hf3fs_connector.py:116
↓ 2 callers
Method
get_deserialize_event_fd
Return the fd signaled on deserialize completion.
lmcache/v1/distributed/serde/async_processor.py:78
↓ 2 callers
Method
get_event_fd
Return the notifier fd that is signaled when new keys are available. Returns: int: The readable file descriptor
lmcache/v1/distributed/storage_controllers/store_controller.py:85
↓ 2 callers
Method
get_eviction_actions
Get the eviction actions to evict objects from cache. Args: expected_ratio (float): A hint indicating approximately what
lmcache/v1/distributed/eviction.py:86
↓ 2 callers
Method
get_fill_rate
(self)
lmcache/v1/utils/bloom_filter.py:95
↓ 2 callers
Method
get_general_items
Return phase-2 (general) items that are not yet set.
lmcache/cli/commands/bench/engine_bench/interactive/state.py:120
↓ 2 callers
Method
get_handlers
Return handler specs for all request types this module serves. Returns: A list of HandlerSpec entries mapping request types to
lmcache/v1/multiprocess/modules/blend.py:336
↓ 2 callers
Method
get_instance
Initialize the HF3FS client
lmcache/v1/storage_backend/connector/hf3fs_connector.py:307
↓ 2 callers
Method
get_instance_by_ip
Get an instance by IP address. Returns first instance if multiple exist.
lmcache/v1/cache_controller/utils.py:461
↓ 2 callers
Method
get_kernel_group_kv_pointers
Returns the KV pointer tensor for the given kernel group. Mirrors :meth:`GPUCacheContext.get_kernel_group_kv_pointers`.
lmcache/v1/platform/cpu/cache_context.py:288
↓ 2 callers
Method
get_keys
Return predefined keys.
tests/v1/cache_controller/conftest.py:132
↓ 2 callers
Method
get_kv_buffer_shape
Returns the KV buffer shape for *logical_num_tokens*.
lmcache/v1/platform/base_cache_context.py:208
↓ 2 callers
Method
get_kv_cache
Wrap GPU tensors in CudaIPCWrapper for IPC communication.
tests/v1/multiprocess/test_cache_server.py:117
↓ 2 callers
Function
get_layout_desc
Get the memory layout description for a specific object group. The returned layout describes the single memory object that backs ``object_gro
lmcache/v1/multiprocess/modules/lmcache_driven_transfer.py:56
↓ 2 callers
Method
get_load_event_fd
(self)
examples/lmc_external_l2_adapter/src/lmc_external_l2_adapter/adapter.py:155
↓ 2 callers
Method
get_load_event_fd
Return the eventfd signaled when load tasks complete.
lmcache/v1/distributed/l2_adapters/raw_block_l2_adapter.py:401
↓ 2 callers
Method
get_memory_indices
Get L1 memory page indices for the given address and size.
lmcache/v1/distributed/l2_adapters/nixl_store_dynamic_l2_adapter.py:266
↓ 2 callers
Method
get_memory_indices
Get memory indices for the given raw address and size.
lmcache/v1/distributed/l2_adapters/nixl_store_l2_adapter.py:391
↓ 2 callers
Method
get_message_count
(self, benchmark: "ZMQControllerBenchmark")
lmcache/tools/controller_benchmark/handlers/evict.py:51
↓ 2 callers
Function
get_methods_implemented_in_class
Get methods that are actually implemented in the class itself. Args: cls: The class to inspect base_class: Optional base clas
tests/v1/utils.py:652
↓ 2 callers
Method
get_num_groups
(self)
lmcache/v1/metadata.py:114
↓ 2 callers
Method
get_paths_info
Return exact metadata entries for the requested paths.
lmcache/v1/storage_backend/connector/hfbucket_connector.py:139
↓ 2 callers
Function
get_remote_metadata_bytes
()
lmcache/v1/protocol.py:93
↓ 2 callers
Method
get_response_class
(self)
lmcache/v1/multiprocess/mq.py:399
↓ 2 callers
Method
get_running_count
Get the count of running threads.
lmcache/v1/periodic_thread.py:489
↓ 2 callers
Method
get_shape
Get the shape of the data for a single layer. Args: num_tokens: The number of tokens in the data. Returns: T
lmcache/v1/gpu_connector/musa_connectors.py:823
↓ 2 callers
Method
get_shape
(self, num_tokens: int)
lmcache/v1/gpu_connector/gpu_connectors.py:1929
↓ 2 callers
Method
get_shape_desc
Return the :class:`PageBufferShapeDesc` for *kernel_group_idx*. Args: kernel_group_idx: 0-based kernel group index. Rais
lmcache/v1/kv_layer_groups.py:449
↓ 2 callers
Method
get_size
(self)
lmcache/v1/memory_management.py:669
↓ 2 callers
Method
get_size
(self)
lmcache/v1/memory_management.py:921
↓ 2 callers
Method
get_slots_per_chunk_in_sw
Returns the number of slots per lmcache chunk for D/H transfer.
lmcache/v1/platform/base_cache_context.py:203
↓ 2 callers
Method
get_tmp_gpu_buffer
Returns the temporary GPU buffer for transfers
lmcache/v1/platform/cuda/cache_context.py:588
↓ 2 callers
Method
get_total_kv_count
Get total count of KV chunks across all workers in this instance.
lmcache/v1/cache_controller/utils.py:331
↓ 2 callers
Method
get_transfer_channel_address
Translate ``(offset, size)`` L1 objects into the transfer-channel-specific addresses. Args: lmcache_addresses: A list of
lmcache/v1/distributed/transfer_channel/abstract.py:139
↓ 2 callers
Method
get_transfer_channel_client
Get or create a client that is connected to the peer identified by ``peer_advertise_url``. Args: peer_advertise_url: The
lmcache/v1/distributed/transfer_channel/abstract.py:102
↓ 2 callers
Method
get_usage_bytes_by_cache_salt
Aggregate ``cache_salt`` byte usage across every L2 adapter. Used by the HTTP quota endpoints to report ``current_usage_gb`` alongsid
lmcache/v1/distributed/storage_manager.py:751
↓ 2 callers
Method
get_worker
Get a specific worker.
lmcache/v1/cache_controller/utils.py:450
↓ 2 callers
Method
get_worker_ids
Get sorted list of worker IDs for an instance.
lmcache/v1/cache_controller/utils.py:478
↓ 2 callers
Function
group_layers_by_identity
Partition layer indices by :data:`LayerGroupIdentity`. This helper is shared by vLLM-side LMCache group inflation and server-side ``KernelGro
lmcache/v1/kv_layer_groups.py:70
↓ 2 callers
Method
handle_init_side_msg
Handle side messages during initialization. :param req: The initialization-related side message received from the peer.
lmcache/v1/transfer_channel/abstract.py:70
↓ 2 callers
Method
handler
(**kw)
tests/v1/test_basic_check.py:126
↓ 2 callers
Method
has_cache_item
Return whether LMCache already contains the encoder cache for hash.
lmcache/integration/vllm/vllm_ec_adapter.py:167
↓ 2 callers
Function
has_interprocess_parameter
(obj)
lmcache/utils.py:71
↓ 2 callers
Method
has_kv
Check if a KV chunk exists in this worker.
lmcache/v1/cache_controller/utils.py:162
↓ 2 callers
Method
has_subscribers
Return True if at least one callback is registered for *event_type*. Use this to skip expensive event construction on the hot path when
lmcache/v1/mp_observability/event_bus.py:156
↓ 2 callers
Method
has_worker_with_ip
Check if any worker in this instance has the specified IP address.
lmcache/v1/cache_controller/utils.py:324
↓ 2 callers
Method
has_workers
Check if instance has any workers.
lmcache/v1/cache_controller/utils.py:282
↓ 2 callers
Function
hash_to_tokens
Deterministically map *chunk_hash* to *n_tokens* token IDs. The mapping is a pure function of ``chunk_hash`` and ``n_tokens`` — the same inpu
lmcache/tools/cache_simulator/gen_bench_dataset.py:116
↓ 2 callers
Method
health
Execute a compress operation with error handling.
lmcache/v1/cache_controller/executor.py:352
↓ 2 callers
Method
hotplug_add_device
Map and activate one additional DAX device. Args: device_path: Path to an existing readable and writable DAX device.
lmcache/v1/distributed/l2_adapters/dax_l2_adapter.py:768
↓ 2 callers
Method
increase_num_scheduled_tokens
(self, num_new_tokens: int)
lmcache/integration/vllm/lmcache_mp_connector.py:243
↓ 2 callers
Method
increase_num_scheduled_tokens
(self, num_new_tokens: int)
lmcache/integration/vllm/lmcache_mp_connector_0201.py:256
↓ 2 callers
Method
increase_num_scheduled_tokens
(self, num_new_tokens: int)
lmcache/integration/vllm/lmcache_mp_connector_0180.py:238
↓ 2 callers
Method
increase_num_stored_blocks
Increase the number of stored blocks for the current request This function will be called when processing the cached requests.
lmcache/integration/vllm/lmcache_mp_connector_0201.py:259
↓ 2 callers
Method
increase_num_stored_blocks
Increase the number of stored blocks for the current request This function will be called when processing the cached requests.
lmcache/integration/vllm/lmcache_mp_connector_0180.py:241
↓ 2 callers
Method
increase_num_stored_tokens
Increase the number of stored tokens for the current request This function will be called when processing the cached requests.
lmcache/integration/vllm/lmcache_mp_connector.py:246
↓ 2 callers
Method
increase_pin_count
(self)
lmcache/v1/distributed/l2_adapters/nixl_store_l2_adapter.py:62
↓ 2 callers
Method
init_chunk_meta
Initialize chunk metadata similar to base_connector.init_chunk_meta()
lmcache/v1/storage_backend/nixl_storage_backend.py:1467
↓ 2 callers
Method
initialize_allocator
( self, config: LMCacheEngineConfig, metadata: LMCacheMetadata )
lmcache/v1/storage_backend/pd_backend.py:296
↓ 2 callers
Function
initialize_gds_context
Set up the process-global :class:`GDSContext` (once, at startup). ``config=None`` leaves it uninitialized (GDS L1 disabled); otherwise the sl
lmcache/v1/gpu_connector/gds_context.py:399
↓ 2 callers
Method
initialize_kvcaches_ptr
Initialize the kvcaches pointers if not already initialized.
lmcache/v1/gpu_connector/hpu_connector.py:240
↓ 2 callers
Function
initialize_nodes
Initialize node configuration from CLI args or supplier URL. Args: node_supplier_url: Optional discovery service URL. When set,
lmcache/lmcache_frontend/app.py:750
↓ 2 callers
Method
invalidate
Invalidate the MemoryObj.
lmcache/v1/memory_management.py:208
↓ 2 callers
Function
is_dp_rank0
(vllm_config: "VllmConfig")
lmcache/integration/vllm/utils.py:418
↓ 2 callers
Method
is_expired
Check if the cache is expired without refreshing it. Args: timeout_override: Optional timeout override in seconds.
lmcache/v1/utils/cache_utils.py:71
↓ 2 callers
Method
is_explicitly_requested
Return True when this profile was selected via env var.
setup_extensions/build_profiles/__init__.py:104
↓ 2 callers
Function
is_hnd
Return ``True`` if the Engine KV Format uses an HND physical layout.
lmcache/v1/gpu_connector/utils.py:99
↓ 2 callers
Function
is_native_musa_kv_transfer_enabled
Return whether LMCache should try optional native MUSA KV transfer.
lmcache/v1/platform/musa/native_kv_transfer.py:37
↓ 2 callers
Function
is_system
(system)
benchmarks/multi_round_qa/data_preprocessing.py:31
↓ 2 callers
Method
isatty
(self)
tests/test_banner.py:24
↓ 2 callers
Method
launch_plugins
Launch all configured plugins.
lmcache/v1/multiprocess/mp_runtime_plugin_launcher.py:103
↓ 2 callers
Method
launch_request
finish_callback: Callable[[Response], None]
benchmarks/multi_round_qa/multi-round-qa.py:169
↓ 2 callers
Method
launch_request
finish_callback: Callable[[Response], None]
benchmarks/rag/rag.py:212
↓ 2 callers
Method
list
List all keys.
tests/v1/storage_backend/test_remote_storage_plugin.py:63
↓ 2 callers
Function
lmcache_get_config
Load the LMCache configuration. Args: config_file: Path to a YAML configuration file, or empty string to build the config fro
lmcache/integration/sglang/utils.py:15
↓ 2 callers
Function
lmcache_memcpy_async_d2h
Helper function to copy memory object allocated by different allocators from GPU buffer. This function is non-blocking and won't do stream sy
lmcache/v1/gpu_connector/gpu_ops.py:54
↓ 2 callers
Function
loadConfig
()
lmcache/lmcache_frontend/static/js/app.js:811
↓ 2 callers
Function
loadLogLevel
()
lmcache/lmcache_frontend/static/js/app.js:744
← previous
next →
2,701–2,800 of 13,371, ranked by callers