MCPcopy Create free account

hub / github.com/bytedance/ABQ-LLM / functions

Functions14,558 in github.com/bytedance/ABQ-LLM

⨍ Functions 14,558◇ Types & classes 4,979
↓ 2 callersFunctioninit_op_cache
(decoder_args, batchxbeam, memory_max_seq_len, decoding_max_seq_len)
fastertransformer/examples/tensorflow/decoder/utils/decoder.py:304
↓ 2 callersMethodinitialize
(self)
fastertransformer/3rdparty/cutlass/tools/library/scripts/pycutlass/src/pycutlass/parser.py:550
↓ 2 callersMethodinitialize
Initializes data structures
fastertransformer/3rdparty/cutlass/test/unit/gemm/device/testbed_trmm_universal.h:239
↓ 2 callersMethodinitialize
Initializes data structures
fastertransformer/3rdparty/cutlass/test/unit/gemm/device/testbed_symm_universal.h:193
↓ 2 callersMethodinitialize
Initializes data structures
fastertransformer/3rdparty/cutlass/test/unit/gemm/device/testbed_rank2k_universal.h:193
↓ 2 callersMethodinitialize
Initializes data structures
fastertransformer/3rdparty/cutlass/test/unit/gemm/device/testbed_sparse.h:164
↓ 2 callersMethodinitialize
Initializes data structures
fastertransformer/3rdparty/cutlass/test/unit/gemm/device/testbed_universal.h:144
↓ 2 callersMethodinitialize
Initializes data structures
fastertransformer/3rdparty/cutlass/test/unit/gemm/device/testbed_rank_k_universal.h:189
↓ 2 callersMethodinitialize
fastertransformer/3rdparty/cutlass/test/unit/conv/device/conv2d_testbed.h:170
↓ 2 callersMethodinitialize
fastertransformer/3rdparty/cutlass/test/unit/conv/device/conv3d_testbed.h:162
↓ 2 callersMethodinitialize
fastertransformer/3rdparty/cutlass/test/unit/conv/device/conv2d_testbed_interleaved.h:164
↓ 2 callersFunctioninitialize_decoding_variables
(decoding_args, batchxbeam)
fastertransformer/examples/tensorflow/decoder/utils/decoding.py:30
↓ 2 callersFunctioninput_fn_builder
Creates an `input_fn` closure to be passed to TPUEstimator.
fastertransformer/examples/tensorflow/bert/tensorflow_bert/bert/run_squad.py:687
↓ 2 callersFunctioninput_fn_builder
Creates an `input_fn` closure to be passed to TPUEstimator.
fastertransformer/examples/tensorflow/bert/tensorflow_bert/bert/run_pretraining.py:324
↓ 2 callersFunctioninput_fn_builder
Creates an `input_fn` closure to be passed to Estimator.
fastertransformer/examples/tensorflow/bert/bert-quantization/run_squad.py:571
↓ 2 callersFunctioninput_fn_builder
Creates an `input_fn` closure to be passed to Estimator.
fastertransformer/examples/tensorflow/bert/bert-quantization/run_pretraining.py:412
↓ 2 callersFunctioninsert_dict
(words, ngrams, pos)
fastertransformer/3rdparty/Megatron-LM/tools/openwebtext/filter_ngrams.py:187
↓ 2 callersMethodinverse
Computes the inverse of a 2-by-2 matrix given the matrix's determinant
fastertransformer/3rdparty/cutlass/include/cutlass/matrix.h:3240
↓ 2 callersFunctioninvokeAddBiasGeluV2
fastertransformer/src/fastertransformer/kernels/activation_kernels.h:167
↓ 2 callersFunctionisFusedMHA
fastertransformer/src/fastertransformer/layers/attention_layers/BaseAttentionLayer.h:115
↓ 2 callersFunctionisPaddedMHA
fastertransformer/src/fastertransformer/layers/attention_layers/BaseAttentionLayer.h:125
↓ 2 callersMethodisValid
fastertransformer/src/fastertransformer/utils/Tensor.h:313
↓ 2 callersMethodis_complex
(self)
fastertransformer/3rdparty/cutlass/tools/library/scripts/conv2d_operation.py:35
↓ 2 callersMethodis_complex
(self)
fastertransformer/3rdparty/cutlass/tools/library/scripts/pycutlass/src/pycutlass/conv2d_operation.py:524
↓ 2 callersMethodis_f16c_supported
fastertransformer/3rdparty/cutlass/include/cutlass/half.h:164
↓ 2 callersFunctionis_pipeline_stage_before_split
Return True if pipeline stage executes encoder block for a model with both encoder and decoder.
fastertransformer/3rdparty/Megatron-LM/megatron/mpu/initialize.py:390
↓ 2 callersFunctionis_start_piece
Check if the current word piece is the starting piece (BERT).
fastertransformer/3rdparty/Megatron-LM/megatron/data/dataset_utils.py:172
↓ 2 callersFunctionis_within_directory
(directory, target)
fastertransformer/examples/pytorch/nemo.py:43
↓ 2 callersFunctionisfinite
fastertransformer/3rdparty/cutlass/include/cutlass/tfloat32.h:197
↓ 2 callersFunctionisfinite
fastertransformer/3rdparty/cutlass/include/cutlass/half.h:515
↓ 2 callersFunctionjoin
fastertransformer/examples/cpp/multi_gpu_gpt/multi_gpu_gpt_async_example.cc:47
↓ 2 callersMethodlast_boxed_only_string
(self, string)
algorithm/lm_eval/tasks/hendrycks_math.py:117
↓ 2 callersFunctionlayer_norm
(input_tensor, name=None)
fastertransformer/examples/tensorflow/bert/utils/bert.py:33
↓ 2 callersMethodlisted_weights
(self)
fastertransformer/examples/pytorch/vit/VisionTransformerINT8WeightLoader.py:115
↓ 2 callersFunctionload
(args)
fastertransformer/3rdparty/Megatron-LM/megatron/fused_kernels/__init__.py:30
↓ 2 callersMethodload
Loads a fragment
fastertransformer/src/fastertransformer/cutlass_extensions/include/cutlass_extensions/epilogue/threadblock/epilogue_tensor_op_int32.h:271
↓ 2 callersMethodload
(self, ckpt_path)
fastertransformer/examples/pytorch/llama2/llamav2.py:255
↓ 2 callersFunctionloadWeightsPtr
fastertransformer/src/fastertransformer/tensorrt_plugin/vit/ViTPlugin.cpp:416
↓ 2 callersFunctionloadWeightsPtrINT8
fastertransformer/src/fastertransformer/tensorrt_plugin/vit/ViTINT8Plugin.cpp:431
↓ 2 callersMethodload_articles
(self)
fastertransformer/examples/pytorch/bert/bert-quantization-sparsity/data/TextSharding.py:46
↓ 2 callersFunctionload_biencoder_checkpoint
selectively load retrieval models for indexing/retrieving from saved checkpoints
fastertransformer/3rdparty/Megatron-LM/megatron/checkpointing.py:452
↓ 2 callersMethodload_from
(self, weights, n_block)
fastertransformer/examples/pytorch/vit/ViT-quantization/vit_int8.py:329
↓ 2 callersMethodload_from
(self, weights, n_block)
fastertransformer/examples/pytorch/vit/ViT-quantization/ViT-pytorch/models/modeling.py:192
↓ 2 callersMethodload_from_file
Populate members from instance saved to file
fastertransformer/3rdparty/Megatron-LM/megatron/data/realm_index.py:50
↓ 2 callersMethodload_from_model
(self, model)
fastertransformer/examples/pytorch/t5/utils/ft_decoding.py:70
↓ 2 callersMethodload_from_model
(self, model)
fastertransformer/examples/pytorch/t5/utils/ft_encoder.py:71
↓ 2 callersFunctionload_the_npz_array
fastertransformer/examples/cpp/xlnet/cnpy.cpp:215
↓ 2 callersFunctionload_vocab
Loads a vocabulary file into a dictionary.
fastertransformer/examples/tensorflow/bert/bert-quantization/tokenization.py:115
↓ 2 callersFunctionload_with_pointer_offset
Load
fastertransformer/3rdparty/cutlass/include/cutlass/epilogue/warp/tile_iterator_volta_tensor_op.h:216
↓ 2 callersFunctionlog2_ceil
fastertransformer/3rdparty/Megatron-LM/megatron/fused_kernels/scaled_upper_triang_masked_softmax.h:65
↓ 2 callersFunctionloss_func
(loss_mask, output_tensor)
fastertransformer/3rdparty/Megatron-LM/pretrain_t5.py:112
↓ 2 callersFunctionloss_func
(output_predictions, labels, output_tensor)
fastertransformer/3rdparty/Megatron-LM/tasks/eval_utils.py:103
↓ 2 callersFunctionlskip
Return pointer to first non-whitespace char in given string. */
fastertransformer/3rdparty/INIReader.h:127
↓ 2 callersMethodmainLoop
engine/mma_any/aq_wmma_kernel.h:120
↓ 2 callersMethodmainLoop
algorithm/ABQ/csrc/mma_any/aq_wmma_kernel.h:115
↓ 2 callersMethodmainLoop
fastertransformer/src/fastertransformer/kernels/abqgemm/mma_any/aq_wmma_kernel.h:101
↓ 2 callersFunctionmake_attention_mask
Returns a 2-dimensional (2-D) attention mask :param source_block: 1-D array :param target_block: 1-D array
fastertransformer/3rdparty/Megatron-LM/megatron/data/ict_dataset.py:12
↓ 2 callersMethodmax_dim_index
Returns the index of the dimension with greatest value
fastertransformer/3rdparty/cutlass/include/cutlass/coord.h:133
↓ 2 callersFunctionmerge_partitions
(merged, partitions, partition_dim, stride)
fastertransformer/3rdparty/Megatron-LM/tools/merge_mp_partitions.py:55
↓ 2 callersFunctionmetric_fn
(per_example_loss, label_ids, logits, is_real_example)
fastertransformer/examples/tensorflow/bert/tensorflow_bert/bert/run_classifier.py:684
↓ 2 callersFunctionmetric_fn
Computes the loss and accuracy of the model.
fastertransformer/examples/tensorflow/bert/bert-quantization/run_pretraining.py:276
↓ 2 callersFunctionmetric_max_over_ground_truths
Compute max metric between prediction and each ground truth.
algorithm/lm_eval/metrics.py:85
↓ 2 callersFunctionmetric_max_over_ground_truths
(metric_fn, prediction, ground_truths)
fastertransformer/examples/tensorflow/bert/tensorflow_bert/squad_evaluate-v1.1.py:60
↓ 2 callersFunctionmetric_max_over_ground_truths
(metric_fn, prediction, ground_truths)
fastertransformer/examples/tensorflow/bert/tensorflow_bert/squad_evaluate_v1_1.py:60
↓ 2 callersFunctionmkdir
(path)
fastertransformer/examples/pytorch/bert/bert-quantization-sparsity/utils.py:58
↓ 2 callersMethodmnk
Obtains a Coord<3> from GemmCoord
fastertransformer/3rdparty/cutlass/include/cutlass/gemm/gemm.h:157
↓ 2 callersFunctionmodel_fn_builder
(bert_config, transformer_model_fn)
fastertransformer/examples/tensorflow/bert/tensorflow_bert/profile_transformer_inference.py:97
↓ 2 callersFunctionnorm
fastertransformer/3rdparty/cutlass/include/cutlass/quaternion.h:409
↓ 2 callersFunctionnormalize_answer
Taken from the official evaluation script for v1.1 of the SQuAD dataset. Lower text and remove punctuation, articles and extra whitespace.
algorithm/lm_eval/tasks/qasper.py:43
↓ 2 callersFunctionnormalize_answer
Lower text and remove punctuation, articles and extra whitespace.
fastertransformer/3rdparty/Megatron-LM/tasks/msdp/metrics.py:18
↓ 2 callersMethodo_strideM
fastertransformer/3rdparty/cutlass/examples/42_fused_multi_head_attention/kernel_forward.h:132
↓ 2 callersFunctionoa_completion
Query OpenAI API for completion. Retry with back-off until they respond
algorithm/lm_eval/models/gpt3.py:38
↓ 2 callersMethodoperand_B_ref
Returns a TensorRef to the B operand
fastertransformer/3rdparty/cutlass/examples/42_fused_multi_head_attention/mma_from_smem.h:217
↓ 2 callersMethodoperator++
fastertransformer/3rdparty/cutlass/examples/42_fused_multi_head_attention/iterators/predicated_tile_access_iterator_residual_last.h:846
↓ 2 callersMethodoperator++
fastertransformer/3rdparty/cutlass/include/cutlass/transform/threadblock/predicated_tile_access_iterator.h:986
↓ 2 callersMethodoperator++
fastertransformer/3rdparty/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator_tensor_op.h:710
↓ 2 callersMethodoutput
Returns Output Tensor
fastertransformer/3rdparty/cutlass/tools/profiler/src/cudnn_helpers.h:576
↓ 2 callersFunctionp
(by,bw,bv)
fastertransformer/3rdparty/cutlass/docs/jquery.js:23
↓ 2 callersFunctionpad_in_time
Helper function to pad a tensor in the time dimension and retain the static depth dimension.
fastertransformer/examples/tensorflow/bert/utils/reducer.py:18
↓ 2 callersFunctionpad_in_time
Helper function to pad a tensor in the time dimension and retain the static depth dimension.
fastertransformer/examples/tensorflow/decoder/utils/reducer.py:18
↓ 2 callersFunctionpad_in_time
Helper function to pad a tensor in the time dimension and retain the static depth dimension.
fastertransformer/examples/tensorflow/common_utils/reducer.py:18
↓ 2 callersFunctionparallel_self_attention
(tensor_model_parallel_size, num_att_heads_per_partition, hidden_size_per_att_head
fastertransformer/3rdparty/Megatron-LM/megatron/mpu/tests/test_layers.py:328
↓ 2 callersFunctionparallel_transformer
(tensor_model_parallel_size, num_att_heads_per_partition, hidden_size_per_att_head, b
fastertransformer/3rdparty/Megatron-LM/megatron/mpu/tests/test_layers.py:413
↓ 2 callersMethodpartial_context
(cls, doc, option)
algorithm/lm_eval/tasks/winogrande.py:66
↓ 2 callersMethodpartial_context
(cls, doc, option)
algorithm/lm_eval/tasks/wsc273.py:106
↓ 2 callersMethodpartial_target
(cls, doc)
algorithm/lm_eval/tasks/winogrande.py:76
↓ 2 callersMethodpartial_target
(cls, doc)
algorithm/lm_eval/tasks/wsc273.py:115
↓ 2 callersFunctionpil_loader
(path: str)
fastertransformer/3rdparty/Megatron-LM/megatron/data/image_folder.py:237
↓ 2 callersFunctionpil_loader
(path)
fastertransformer/examples/pytorch/swin/Swin-Transformer-Quantization/SwinTransformer/data/cached_image_folder.py:178
↓ 2 callersMethodplan
(self, arguments)
fastertransformer/3rdparty/cutlass/tools/library/scripts/pycutlass/src/pycutlass/gemm_operation.py:726
↓ 2 callersFunctionpositionwise_ffn
Position-wise Feed-forward Network.
fastertransformer/examples/tensorflow/xlnet/modeling.py:65
↓ 2 callersFunctionpr_re_fbeta
Uses a confusion matrix to compute precision, recall and fbeta
fastertransformer/examples/tensorflow/bert/bert-quantization/tf_metrics.py:145
↓ 2 callersFunctionprepare_input
(question, passage_text, seq_len, batch_size, model_dir, data_type)
fastertransformer/examples/pytorch/longformer/longformer_qa.py:58
↓ 2 callersMethodpreprocess
(cls, text)
algorithm/lm_eval/tasks/hellaswag.py:62
↓ 2 callersFunctionprint_progress
(prefix, start_time, urls_counter, domain_blacklist_counter, extention_b
fastertransformer/3rdparty/Megatron-LM/tools/openwebtext/blacklist_urls.py:233
↓ 2 callersFunctionprint_progress
(prefix, start_time, num_docs, num_fixed_text, num_non_english_docs, chars_non_english_docs
fastertransformer/3rdparty/Megatron-LM/tools/openwebtext/cleanup_dataset.py:30
↓ 2 callersFunctionprint_quant_summary
Print summary of all quantizer modules in the model.
fastertransformer/examples/pytorch/vit/ViT-quantization/quant_utils.py:216
↓ 2 callersFunctionprint_quant_summary
Print summary of all quantizer modules in the model.
fastertransformer/examples/pytorch/bert/bert-quantization-sparsity/quant_utils.py:216
↓ 2 callersFunctionprint_usage
fastertransformer/3rdparty/cutlass/examples/03_visualize_layout/visualize_layout.cpp:52
↓ 2 callersFunctionprocess_batch
Process batch and produce inputs for the model.
fastertransformer/3rdparty/Megatron-LM/tasks/finetune_utils.py:38
← previousnext →1,701–1,800 of 14,558, ranked by callers