Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/bytedance/ABQ-LLM
/ functions
Functions
14,558 in github.com/bytedance/ABQ-LLM
⨍
Functions
14,558
◇
Types & classes
4,979
↓ 2 callers
Function
flatten_adapter
Flatten adapter weights.
fastertransformer/examples/pytorch/t5/utils/nemo_t5_ckpt_convert.py:655
↓ 2 callers
Method
float32
fastertransformer/3rdparty/cutlass/tools/library/scripts/pycutlass/src/cpp/include/types.h:73
↓ 2 callers
Function
float_to_half
fastertransformer/src/fastertransformer/kernels/decoder_masked_multihead_attention_utils.h:288
↓ 2 callers
Function
form_ngrams
(sequence, n)
algorithm/lm_eval/decontamination/janitor.py:22
↓ 2 callers
Method
format_answer
(cls, query, entity)
algorithm/lm_eval/tasks/superglue.py:321
↓ 2 callers
Method
forward
Run a forward pass for each of the models and return the respective embeddings.
fastertransformer/3rdparty/Megatron-LM/megatron/model/realm_model.py:66
↓ 2 callers
Method
forward
(ctx, inputs, scale)
fastertransformer/3rdparty/Megatron-LM/megatron/model/fused_softmax.py:92
↓ 2 callers
Method
forward
fastertransformer/src/fastertransformer/models/bart/BartDecoder.cc:291
↓ 2 callers
Method
forward
fastertransformer/src/fastertransformer/models/longformer/LongformerEncoder.cc:140
↓ 2 callers
Method
forward
fastertransformer/src/fastertransformer/models/t5/T5Decoding.cc:355
↓ 2 callers
Method
forward
(self, input_ids, seq_len, remove_padding=True)
fastertransformer/examples/tensorflow/deberta/utils/ft_deberta.py:160
↓ 2 callers
Method
forward
(self, *args, **kwargs)
fastertransformer/examples/pytorch/longformer/model.py:83
↓ 2 callers
Function
forward_step_helper
Helper method to run forward step with model split into chunks (run set_virtual_pipeline_model_parallel_rank() before calling forward_
fastertransformer/3rdparty/Megatron-LM/megatron/schedules.py:294
↓ 2 callers
Method
from_json_file
Constructs a `BertConfig` from a json file of parameters.
fastertransformer/examples/tensorflow/bert/tensorflow_bert/my_modeling.py:94
↓ 2 callers
Function
ftNcclCommInitRank
fastertransformer/src/fastertransformer/utils/nccl_utils.cc:282
↓ 2 callers
Function
fuse_qkv
Adjust quantization ranges to match an implementation where the QKV projections are implemented with a single GEMM. Force the weight and output sc
fastertransformer/examples/pytorch/vit/ViT-quantization/quant_utils.py:189
↓ 2 callers
Function
fuse_qkv
Adjust quantization ranges to match an implementation where the QKV projections are implemented with a single GEMM. Force the weight and output sc
fastertransformer/examples/pytorch/bert/bert-quantization-sparsity/quant_utils.py:189
↓ 2 callers
Function
fuse_qkv
Adjust quantization ranges to match an implementation where the QKV projections are implemented with a single GEMM. Force the weight and output sc
fastertransformer/examples/pytorch/swin/Swin-Transformer-Quantization/quant_utils.py:160
↓ 2 callers
Function
gather_nd
(params, indices)
fastertransformer/examples/pytorch/decoding/utils/decoding.py:141
↓ 2 callers
Method
gbytes_per_sec
memory bandwidth in units of GiB/s
fastertransformer/3rdparty/cutlass/tools/profiler/src/performance_result.h:116
↓ 2 callers
Function
gemm_A_in_regs
fastertransformer/src/fastertransformer/kernels/llama/flash_attention2/utils.h:85
↓ 2 callers
Function
gen_tensor_dataset
(features)
fastertransformer/examples/pytorch/bert/bert-quantization-sparsity/run_glue.py:406
↓ 2 callers
Function
generate_encoder_igemm_config
fastertransformer/src/fastertransformer/utils/gemm_test/encoder_igemm_func.cc:1043
↓ 2 callers
Function
generate_encoder_result
(batch_size, max_seq_len, memory_hidden_dim,
fastertransformer/examples/tensorflow/decoder/utils/decoding.py:53
↓ 2 callers
Method
get
(self)
fastertransformer/3rdparty/cutlass/tools/library/scripts/pycutlass/src/pycutlass/compiler.py:82
↓ 2 callers
Method
get
fastertransformer/3rdparty/cutlass/include/cutlass/subbyte_reference.h:587
↓ 2 callers
Method
getHiddenUnits
fastertransformer/src/fastertransformer/models/multi_gpu_gpt/ParallelGpt.cc:1763
↓ 2 callers
Method
getNumHeads
fastertransformer/src/fastertransformer/models/t5/T5Encoder.h:167
↓ 2 callers
Method
getNumIA3Tasks
fastertransformer/src/fastertransformer/models/t5/T5EncoderWeight.h:66
↓ 2 callers
Method
getNumLayers
fastertransformer/src/fastertransformer/models/t5/T5Encoder.h:162
↓ 2 callers
Function
getOutputSize
get the output size of pooling */
fastertransformer/3rdparty/cutlass/tools/util/include/cutlass/util/device_nhwc_pooling.h:65
↓ 2 callers
Method
getParams
fastertransformer/3rdparty/cutlass/include/cutlass/conv/threadblock/conv2d_tile_iterator.h:102
↓ 2 callers
Function
getSmemPtr
device function to convert shared memory address into unsigned format
engine/common/memory.h:21
↓ 2 callers
Function
getSmemPtr
device function to convert shared memory address into unsigned format
algorithm/ABQ/csrc/common/memory.h:21
↓ 2 callers
Function
getSmemPtr
device function to convert shared memory address into unsigned format
fastertransformer/src/fastertransformer/kernels/abqgemm/common/memory.h:21
↓ 2 callers
Method
getSparseMatrixSize
fastertransformer/src/fastertransformer/utils/cublasMMWrapper.cc:745
↓ 2 callers
Method
getStep
fastertransformer/src/fastertransformer/models/multi_gpu_gpt/ParallelGpt.cc:1775
↓ 2 callers
Function
get_adlr_autoresume
ADLR autoresume object. It can be None so no need to check if it is initialized.
fastertransformer/3rdparty/Megatron-LM/megatron/global_vars.py:69
↓ 2 callers
Method
get_all_encoder_layers
(self)
fastertransformer/examples/tensorflow/bert/tensorflow_bert/bert/modeling.py:246
↓ 2 callers
Function
get_all_gpu_free_memory
()
algorithm/parallel_utils.py:94
↓ 2 callers
Function
get_answer_text
(example, feature, pred, args)
fastertransformer/examples/pytorch/bert/bert-quantization-sparsity/run_squad.py:524
↓ 2 callers
Function
get_args
()
fastertransformer/examples/pytorch/gpt/utils/huggingface_bloom_convert.py:55
↓ 2 callers
Function
get_b16_config
Returns the ViT-B/16 configuration.
fastertransformer/examples/pytorch/vit/ViT-quantization/ViT-pytorch/models/configs.py:34
↓ 2 callers
Function
get_c4
(nsamples, seed, seqlen, model)
algorithm/datautils.py:73
↓ 2 callers
Function
get_calibrators
Prepare collector and relevant tensors for calibration and return a list of calibrators. Args: collection_name_prefix: A string. Determine the
fastertransformer/examples/tensorflow/bert/bert-quantization/ft-tensorflow-quantization/ft_tensorflow_quantization/python/calib/calibrator.py:299
↓ 2 callers
Function
get_complex_from_real
(real_type)
fastertransformer/3rdparty/cutlass/tools/library/scripts/pycutlass/src/pycutlass/library.py:229
↓ 2 callers
Function
get_cudnn_conv_mode
Maps a CUTLASS conv mode to a cuDNN cudnnConvolutionMode_t
fastertransformer/3rdparty/cutlass/tools/profiler/src/cudnn_helpers.cpp:81
↓ 2 callers
Function
get_data_parallel_group
Get the data parallel group the caller rank belongs to.
fastertransformer/3rdparty/Megatron-LM/megatron/mpu/initialize.py:244
↓ 2 callers
Function
get_datasets_weights_and_num_samples
(data_prefix, train_valid_test_num_samples)
fastertransformer/3rdparty/Megatron-LM/megatron/data/dataset_utils.py:44
↓ 2 callers
Method
get_dev_examples
Gets a collection of `InputExample`s for the dev set.
fastertransformer/examples/tensorflow/bert/tensorflow_bert/bert/run_classifier.py:184
↓ 2 callers
Method
get_device_workspace_size
(self, arguments)
fastertransformer/3rdparty/cutlass/tools/library/scripts/pycutlass/src/pycutlass/gemm_operation.py:550
↓ 2 callers
Function
get_file_name
(idx_moe, idx_expert, idx_rank)
fastertransformer/examples/pytorch/gpt/utils/megatron_gpt_moe_ckpt_convert.py:345
↓ 2 callers
Method
get_grid_shape
Computes the grid shape
fastertransformer/3rdparty/cutlass/include/cutlass/gemm/device/gemm_universal_base.h:190
↓ 2 callers
Method
get_knowledge_distillation_loss
(self, output_student, output_teacher)
fastertransformer/examples/pytorch/bert/bert-quantization-sparsity/run_squad.py:71
↓ 2 callers
Method
get_lr
(self)
fastertransformer/examples/pytorch/bert/bert-quantization-sparsity/schedulers.py:126
↓ 2 callers
Method
get_mask
Gets the mask
fastertransformer/3rdparty/cutlass/include/cutlass/transform/threadblock/predicated_tile_iterator_triangular_matrix.h:770
↓ 2 callers
Method
get_mask
Gets the mask
fastertransformer/3rdparty/cutlass/include/cutlass/transform/threadblock/predicated_tile_iterator_2dthreadtile.h:751
↓ 2 callers
Method
get_mask
< Sets the mask
fastertransformer/3rdparty/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_v2.h:444
↓ 2 callers
Function
get_model_files
List all model files that you want to load and convert # Args. model_name: name(like `bigscience/bloom`) or local directory of the model
fastertransformer/examples/pytorch/gpt/utils/huggingface_bloom_convert.py:335
↓ 2 callers
Function
get_model_provider
(only_query_model=False, only_context_model=False, biencoder_shared_query_context_model=False)
fastertransformer/3rdparty/Megatron-LM/megatron/model/biencoder_model.py:18
↓ 2 callers
Function
get_op_cache_config
(size_per_head, dtype)
fastertransformer/examples/tensorflow/decoder/utils/decoder.py:21
↓ 2 callers
Function
get_open_retrieval_wiki_dataset
()
fastertransformer/3rdparty/Megatron-LM/megatron/data/orqa_wiki_dataset.py:28
↓ 2 callers
Method
get_original
(self, newarr)
algorithm/models/models_utils.py:557
↓ 2 callers
Function
get_overlaps_dump_path
(task_name, task_set, ngrams_n_size, limit)
algorithm/lm_eval/decontamination/decontaminate.py:49
↓ 2 callers
Function
get_pairs
Return set of symbol pairs in a word. Word is represented as tuple of symbols (symbols being variable-length strings).
fastertransformer/3rdparty/Megatron-LM/megatron/tokenizer/gpt2_tokenization.py:79
↓ 2 callers
Function
get_pairs
Return set of symbol pairs in a word. Word is represented as tuple of symbols (symbols being variable-length strings).
fastertransformer/examples/tensorflow/gpt/utils/gpt_token_encoder.py:68
↓ 2 callers
Function
get_pairs
Return set of symbol pairs in a word. Word is represented as tuple of symbols (symbols being variable-length strings).
fastertransformer/examples/pytorch/gpt/utils/gpt_token_encoder.py:68
↓ 2 callers
Method
get_parameters
(self)
fastertransformer/3rdparty/Megatron-LM/megatron/optimizer/optimizer.py:87
↓ 2 callers
Function
get_pipeline_model_parallel_group
Get the pipeline model parallel group the caller rank belongs to.
fastertransformer/3rdparty/Megatron-LM/megatron/mpu/initialize.py:237
↓ 2 callers
Function
get_ptb
(nsamples, seed, seqlen, model)
algorithm/datautils.py:51
↓ 2 callers
Method
get_reordered
(self)
algorithm/models/models_utils.py:554
↓ 2 callers
Function
get_samples_mapping
Get a list that maps a sample index to a starting sentence index, end sentence index, and length
fastertransformer/3rdparty/Megatron-LM/megatron/data/dataset_utils.py:633
↓ 2 callers
Method
get_sequence_output
Gets final hidden layer of encoder. Returns: float Tensor of shape [batch_size, seq_length, hidden_size] corresponding to the final h
fastertransformer/examples/tensorflow/bert/tensorflow_bert/my_modeling.py:243
↓ 2 callers
Method
get_sequence_output
Gets final hidden layer of encoder. Returns: float Tensor of shape [batch_size, seq_length, hidden_size] corresponding to the final h
fastertransformer/examples/tensorflow/bert/bert-quantization/modeling.py:253
↓ 2 callers
Method
get_start_end_idx
(self)
fastertransformer/3rdparty/Megatron-LM/megatron/data/data_samplers.py:89
↓ 2 callers
Function
get_strided_dgrad_tile_m
////////////////////////////////////////////////////////////////////////////////////////////
fastertransformer/3rdparty/cutlass/include/cutlass/conv/threadblock/threadblock_swizzle.h:53
↓ 2 callers
Function
get_target_sample_len
fastertransformer/3rdparty/Megatron-LM/megatron/data/helpers.cpp:188
↓ 2 callers
Function
get_tensor_model_parallel_src_rank
Calculate the global rank corresponding to the first local rank in the tensor model parallel group.
fastertransformer/3rdparty/Megatron-LM/megatron/mpu/initialize.py:447
↓ 2 callers
Function
get_tensor_shapes
(rank, model_type)
fastertransformer/3rdparty/Megatron-LM/megatron/schedules.py:480
↓ 2 callers
Method
get_tokenizer_file_path
(self, tokenizer_key, file_key, default_filename_pattern)
fastertransformer/examples/pytorch/nemo.py:178
↓ 2 callers
Method
get_train_examples
Gets a collection of `InputExample`s for the train set.
fastertransformer/examples/tensorflow/bert/tensorflow_bert/bert/run_classifier.py:180
↓ 2 callers
Function
get_train_valid_test_split_
Get dataset splits from comma or '/' separated string list.
fastertransformer/3rdparty/Megatron-LM/megatron/data/dataset_utils.py:606
↓ 2 callers
Function
get_valid_prelim_predictions
(start_indices, end_indices, feature, result, args)
fastertransformer/examples/pytorch/bert/bert-quantization-sparsity/run_squad.py:543
↓ 2 callers
Function
get_virtual_pipeline_model_parallel_rank
Return the virtual pipeline-parallel rank.
fastertransformer/3rdparty/Megatron-LM/megatron/mpu/initialize.py:429
↓ 2 callers
Function
get_virtual_pipeline_model_parallel_world_size
Return the virtual pipeline-parallel world size.
fastertransformer/3rdparty/Megatron-LM/megatron/mpu/initialize.py:441
↓ 2 callers
Function
get_weight_data_type
(data_type)
fastertransformer/examples/pytorch/t5/utils/megatron_t5_ckpt_convert.py:95
↓ 2 callers
Function
get_wikitext2
(nsamples, seed, seqlen, model)
algorithm/datautils.py:30
↓ 2 callers
Function
get_words
(text)
fastertransformer/3rdparty/Megatron-LM/tools/openwebtext/filter_ngrams.py:33
↓ 2 callers
Method
get_workspace_size
fastertransformer/src/fastertransformer/kernels/llama/flash_attention2/flash_api.cpp:155
↓ 2 callers
Method
gflops_per_sec
Math throughput in units of GFLOP/s
fastertransformer/3rdparty/cutlass/tools/profiler/src/performance_result.h:111
↓ 2 callers
Method
grid_tiled_shape
(self, problem_size)
fastertransformer/3rdparty/cutlass/tools/library/scripts/rt.py:473
↓ 2 callers
Function
has_file_allowed_extension
Checks if a file is an allowed extension. Args: filename (string): path to a file extensions (tuple of strings): extensions to con
fastertransformer/3rdparty/Megatron-LM/megatron/data/image_folder.py:43
↓ 2 callers
Method
has_test_docs
Whether the task has a test set
algorithm/lm_eval/base.py:464
↓ 2 callers
Function
hmma_fp32
fastertransformer/src/fastertransformer/kernels/decoder_masked_multihead_attention/decoder_masked_multihead_attention_template.hpp:632
↓ 2 callers
Function
hmma_fp32
fastertransformer/src/fastertransformer/kernels/llama/decoder_masked_groupedquery_attention/decoder_masked_groupedquery_attention_template.hpp:632
↓ 2 callers
Method
host_data_ptr_offset
Gets pointer to host data with a pointer offset
fastertransformer/3rdparty/cutlass/tools/util/include/cutlass/util/host_tensor.h:247
↓ 2 callers
Function
i
()
fastertransformer/3rdparty/cutlass/docs/jquery.js:16
↓ 2 callers
Function
ini_parse_file
See documentation in header file. */
fastertransformer/3rdparty/INIReader.h:275
↓ 2 callers
Function
init_graph
()
fastertransformer/examples/tensorflow/ckpt_type_convert.py:28
↓ 2 callers
Function
init_graph
()
fastertransformer/examples/tensorflow/bert/tensorflow_bert/ckpt_type_convert.py:28
← previous
next →
1,601–1,700 of 14,558, ranked by callers