MCPcopy Create free account

hub / github.com/NVIDIA-Merlin/HugeCTR / functions

Functions3,072 in github.com/NVIDIA-Merlin/HugeCTR

⨍ Functions 3,072◇ Types & classes 981↳ Endpoints 3
↓ 1 callersMethodget_dense_variables
(self)
sparse_operation_kit/example/sok_train_hps_inference/sok_train.py:246
↓ 1 callersFunctionget_dev
gpu_cache/include/nv_util.h:67
↓ 1 callersFunctionget_file_size_in_bytes
(filename)
sparse_operation_kit/documents/tutorials/DLRM/bin2csv.py:24
↓ 1 callersFunctionget_file_size_in_bytes
(filename)
test/sok_perf_test/dlrm/bin2csv.py:24
↓ 1 callersMethodget_file_type
HugeCTR/include/data_readers/file_list.hpp:46
↓ 1 callersMethodget_forward_results
only used for results check * Get the forward() results from GPUs and copy them to the host pointer * embedding_feature. This function is only used
HugeCTR/include/embeddings/distributed_slot_sparse_embedding_hash.hpp:316
↓ 1 callersMethodget_forward_results
only used for results check * Get the forward() results from GPUs and copy them to the host pointer * embedding_feature. This function is only used
HugeCTR/include/embeddings/localized_slot_sparse_embedding_hash.hpp:443
↓ 1 callersMethodget_global_gpu_count
sparse_operation_kit/kit_src/lookup/impl/core_impl/tf_backend.hpp:66
↓ 1 callersMethodget_grouped_embedding_tables
HugeCTR/include/embeddings/embedding_collection.hpp:394
↓ 1 callersMethodget_host_data
HugeCTR/include/data_readers/multi_hot/detail/data_reader_impl.hpp:68
↓ 1 callersFunctionget_input_from_json
HugeCTR/src/pybind/add_input.cpp:29
↓ 1 callersFunctionget_input_tensors_and_output_names
HugeCTR/src/pybind/add_dense_layer_helpers.cpp:103
↓ 1 callersMethodget_insert
(self, flatten_ids, length)
sparse_operation_kit/documents/tutorials/DenseDemo/models.py:94
↓ 1 callersMethodget_insert
(self, flatten_ids, length)
test/sok_perf_test/demo/model.py:95
↓ 1 callersMethodget_intra_model_comm_buffer
HugeCTR/embedding/hier_model_parallel_embedding.hpp:110
↓ 1 callersMethodget_intra_reduction_buffer
HugeCTR/embedding/hier_model_parallel_embedding.hpp:112
↓ 1 callersFunctionget_learning_rate_scheduler__
HugeCTR/src/parsers/learning_rate_scheduler_parser.cpp:23
↓ 1 callersMethodget_loaded_data
* @brief Get the loaded data * * @return ptr to the data */
HugeCTR/include/io/file_loader.hpp:70
↓ 1 callersMethodget_local_id
* Get device id according to global id. * @param global_id global id. * @return device_id, if cannot find this device in current node return -1.
HugeCTR/include/device_map.hpp:148
↓ 1 callersFunctionget_logger
(level)
samples/dlrm/preprocessing/convert_to_raw.py:162
↓ 1 callersFunctionget_lookup_id_table_id
HugeCTR/embedding/common.cpp:486
↓ 1 callersMethodget_max_vocabulary_size
HugeCTR/include/embeddings/localized_slot_sparse_embedding_hash.hpp:434
↓ 1 callersMethodget_no_trained_params_in_string
HugeCTR/src/core23_network.cpp:147
↓ 1 callersMethodget_num_row_groups
HugeCTR/src/data_readers/file_source_parquet.cpp:208
↓ 1 callersMethodget_one_read_group
HugeCTR/src/data_readers/row_group_reading_thread.cpp:168
↓ 1 callersMethodget_opt_states_size_in_byte
HugeCTR/include/core23_network.hpp:91
↓ 1 callersMethodget_parallel
HugeCTR/embedding_storage/weight_io/data_info.hpp:108
↓ 1 callersMethodget_params_num
HugeCTR/include/pybind/model.hpp:452
↓ 1 callersMethodget_parquet_metadata
filename is xxxx/_metadata.json
HugeCTR/src/data_readers/metadata.cpp:39
↓ 1 callersMethodget_pid
* Get process id from global id, if not find return -1. */
HugeCTR/include/device_map.hpp:182
↓ 1 callersMethodget_ptr
HugeCTR/include/core23_wrapper.hpp:28
↓ 1 callersMethodget_rank
HugeCTR/core23/logger.hpp:276
↓ 1 callersFunctionget_ref_grad
test/utest/loss/loss_with_regularizer_test.cpp:62
↓ 1 callersFunctionget_ref_term
test/utest/loss/loss_with_regularizer_test.cpp:35
↓ 1 callersFunctionget_region_from_url
HugeCTR/include/io/s3_utils.hpp:130
↓ 1 callersMethodget_replica_uniform_seed
HugeCTR/include/cpu_resource.hpp:36
↓ 1 callersMethodget_row_group
HugeCTR/src/data_readers/file_source_parquet.cpp:207
↓ 1 callersMethodget_rows_file_offset
HugeCTR/include/data_readers/metadata.hpp:82
↓ 1 callersMethodget_rterm
* Return the calculated regularization term */
HugeCTR/include/regularizer.hpp:69
↓ 1 callersFunctionget_sec_accum
test/utest/core23_layer_test/interaction_layer_test.cpp:59
↓ 1 callersFunctionget_sorted_table_ids
HugeCTR/embedding/common.cpp:511
↓ 1 callersFunctionget_sparse_embedding_from_json
HugeCTR/src/pybind/add_sparse_embedding.cpp:29
↓ 1 callersMethodget_stream_name
HugeCTR/src/pipeline.cpp:51
↓ 1 callersFunctionget_table_id_to_ev_size
HugeCTR/embedding/common.cpp:528
↓ 1 callersMethodget_table_keynum
HugeCTR/embedding_storage/weight_io/data_info.hpp:80
↓ 1 callersFunctionget_table_param_list_io
test/utest/embedding_collection/test_embedding_collection_load_dump.cpp:238
↓ 1 callersFunctionget_tensor_from_entities
HugeCTR/src/pybind/add_dense_layer_helpers.cpp:93
↓ 1 callersMethodget_tensor_info_by_name
HugeCTR/src/pybind/model.cpp:1435
↓ 1 callersMethodget_total_inflight_batches
HugeCTR/src/data_readers/multi_hot/detail/data_reader_impl.cpp:227
↓ 1 callersMethodget_universal_batch_size
HugeCTR/include/embedding.hpp:90
↓ 1 callersFunctionget_wgrad_unique_ratio
HugeCTR/embedding/common.cpp:866
↓ 1 callersFunctionglobal_dump_emb_table
(num_gpus, global_embedding_sharding_param_list, global_emb_tables)
notebooks/prototype_embedding_collection/embedding_table.py:103
↓ 1 callersMethodglobal_gpu_id
(self)
sparse_operation_kit/sparse_operation_kit/communication.py:141
↓ 1 callersFunctionglobal_init_key
(num_gpus, global_embedding_sharding_param_list, global_emb_tables, embedding_collection_param, init_keys)
notebooks/prototype_embedding_collection/embedding_table.py:69
↓ 1 callersMethodgreedy_plan
This is a heuristic based on greedy policy. The shard is placed to the bucket with the lowest hotness cost
samples/dlrm/sharding/planner.py:261
↓ 1 callersMethodhas_buffer
HugeCTR/core23/buffer_channel.hpp:52
↓ 1 callersMethodhash
HugeCTR/embedding/operators/unique_op.hpp:80
↓ 1 callersFunctionhctr_print_func
HugeCTR/include/common.hpp:209
↓ 1 callersFunctionhdfs_configs_test
test/utest/io/hdfs_backend_test.cpp:28
↓ 1 callersFunctionhugectr2onnx_bst_test
( batch_size, num_batches, data_file, graph_config, dense_model, sparse_models, on
test/onnx_converter_test/hugectr2onnx_bst_test.py:24
↓ 1 callersFunctionhugectr2onnx_layer_type_test
()
test/onnx_converter_test/layer_type_test.py:21
↓ 1 callersFunctionhugectr2onnx_wdl_test
( batch_size, num_batches, data_file, graph_config, dense_model, sparse_models, on
test/onnx_converter_test/hugectr2onnx_wdl_test.py:24
↓ 1 callersFunctionid_in_rank
()
sparse_operation_kit/sparse_operation_kit/communication.py:210
↓ 1 callersMethodid_in_rank
(self)
sparse_operation_kit/sparse_operation_kit/communication.py:135
↓ 1 callersMethodinc_accomplished_worker
HugeCTR/src/data_readers/row_group_reading_thread.cpp:53
↓ 1 callersMethodindex_calculation_per_gpu
(self, key, bucket_range)
notebooks/prototype_embedding_collection/embedding.py:216
↓ 1 callersFunctioninference_with_saved_model
(args)
sparse_operation_kit/example/sok_train_hps_inference/hps_use_sok_weight_inference.py:177
↓ 1 callersMethodinit
( self, keys: typing.List[int], num_keys: int, offset: typing.List[int],
notebooks/prototype_embedding_collection/embedding_table.py:56
↓ 1 callersMethodinit_ar_comm
HugeCTR/src/exchange_wgrad.cpp:33
↓ 1 callersFunctioninit_data_cpu
test/utest/core23_layer_test/mlp_test.cpp:191
↓ 1 callersFunctioninit_exchange_wgrad
HugeCTR/src/pybind/model.cpp:291
↓ 1 callersFunctioninit_learning_rate_scheduler
HugeCTR/src/pybind/model.cpp:278
↓ 1 callersMethodinit_model_comm_buffer
(self)
notebooks/prototype_embedding_collection/embedding.py:185
↓ 1 callersMethodinit_network_comm_buffer
(self)
notebooks/prototype_embedding_collection/embedding.py:190
↓ 1 callersMethodinitialize
HugeCTR/include/general_buffer2.hpp:144
↓ 1 callersMethodinitialize_max_vocabulary_size_per_gpu
HugeCTR/src/pybind/model.cpp:186
↓ 1 callersFunctionint_to_string
(shard_matrix_int, shard_strategy_int)
samples/dlrm/sharding/generate_plan.py:42
↓ 1 callersFunctionint_to_string
( shard_matrix_int, shard_strategy_int, unique_table_ids_int, reduction_table_ids_int )
benchmarks/embedding_collection/hugectr/sharding/generate_plan.py:33
↓ 1 callersFunctionisVariable
(variable)
sparse_operation_kit/sparse_operation_kit/lookup.py:186
↓ 1 callersMethodis_master_process
HugeCTR/include/resource_managers/resource_manager_core.hpp:89
↓ 1 callersFunctioniter_arrays
(fname, array_ndim=3, dim_dtype=np.int32, array_dtype=np.float32)
test/utest/core23_layer_test/python_concat.py:32
↓ 1 callersMethodkey_map
(self, indices)
sparse_operation_kit/sparse_operation_kit/dynamic_variable.py:287
↓ 1 callersMethodkey_num_per_table
HugeCTR/embedding_storage/dynamic_embedding_cpu.hpp:433
↓ 1 callersFunctionload
Abbreviated as ``sok.load``. Load the embedding tables from sok weight folder. The sok table name must be the same with the weight file
sparse_operation_kit/sparse_operation_kit/dump_load.py:1261
↓ 1 callersMethodload_layer
(self)
onnx_converter/hugectr2onnx/hugectr_loader.py:171
↓ 1 callersFunctionload_per_table
(var, optimizer, path)
sparse_operation_kit/sparse_operation_kit/dump_load.py:1083
↓ 1 callersFunctionload_table
(path, load_vars, optimizer)
sparse_operation_kit/sparse_operation_kit/dump_load.py:1095
↓ 1 callersFunctionload_table_to_filesystem_dynamic
(var, optimizer, path)
sparse_operation_kit/sparse_operation_kit/dump_load.py:970
↓ 1 callersFunctionlookup_sparse_impl
(params, sp_ids, sp_weights=None, combiners=None, use_filter=False)
sparse_operation_kit/sparse_operation_kit/lookup.py:425
↓ 1 callersFunctionmain
(args)
sparse_operation_kit/documents/tutorials/DenseDemo/run_sok_horovod.py:30
↓ 1 callersFunctionmain
(args)
sparse_operation_kit/documents/tutorials/DenseDemo/run_sok_MirroredStrategy.py:28
↓ 1 callersFunctionmain
(args, task_id)
sparse_operation_kit/documents/tutorials/DenseDemo/run_sok_MultiWorker_mpi.py:30
↓ 1 callersFunctionmain
(args, task_id)
sparse_operation_kit/documents/tutorials/DenseDemo/run_tf.py:39
↓ 1 callersFunctionmain
()
sparse_operation_kit/documents/tutorials/MixedPrecision/amp_tf2.py:27
↓ 1 callersFunctionmain
()
sparse_operation_kit/documents/tutorials/MixedPrecision/amp_tf1.py:28
↓ 1 callersFunctionmain
(args)
sparse_operation_kit/documents/tutorials/DLRM/main.py:54
↓ 1 callersFunctionmain
()
samples/dlrm/preprocessing/convert_to_raw.py:227
↓ 1 callersFunctionmain
(args)
test/pybind_test/model_test.py:1236
↓ 1 callersFunctionmain
(args)
test/sok_perf_test/dlrm/main.py:56
↓ 1 callersFunctionmain
(filename, split_num, save_prefix)
test/sok_perf_test/demo/split_data.py:22
← previousnext →801–900 of 3,072, ranked by callers