Functions — github.com/NVIDIA/FasterTransformer

CodeHub MCPcopy Create free account

hub / github.com/NVIDIA/FasterTransformer / functions

Functions4,228 in github.com/NVIDIA/FasterTransformer

⨍ Functions 4,228◇ Types & classes 1,417

↓ 2 callersMethod_gold_score

(self, batch, memory_bank, src_lengths, src_vocabs, use_src_map, enc_states, batch_size, s

examples/pytorch/decoding/utils/translator.py:252

↓ 2 callersMethod_log

examples/pytorch/decoding/utils/translator.py:246

↓ 2 callersMethod_map_int8

examples/pytorch/gpt/utils/gpt.py:235

↓ 2 callersFunction_pil_interp

examples/pytorch/vit/ViT-quantization/data.py:32

↓ 2 callersFunction_prepare_saved_dir

examples/pytorch/gpt/utils/nemo_ckpt_convert.py:394

↓ 2 callersMethod_report_score

(self, name, score_total, words_total)

examples/pytorch/decoding/utils/translator.py:670

↓ 2 callersMethod_run_encoder

examples/pytorch/decoding/utils/translator.py:494

↓ 2 callersFunction_split

(src_key, dst_filename_fn)

examples/pytorch/t5/utils/nemo_t5_ckpt_convert.py:798

↓ 2 callersFunction_strip_spaces

examples/tensorflow/bert/bert-quantization/run_squad.py:844

↓ 2 callersFunction_strip_spaces

examples/pytorch/bert/bert-quantization-sparsity/run_squad.py:614

↓ 2 callersFunctionacc_and_f1

(preds, labels)

examples/pytorch/bert/bert-quantization-sparsity/run_glue.py:106

↓ 2 callersFunctionaddScope

src/fastertransformer/utils/nvtx_utils.cc:29

↓ 2 callersMethodadd_args_group

(parser: argparse.ArgumentParser)

examples/pytorch/gpt/utils/bloom.py:80

↓ 2 callersMethodadvance

src/fastertransformer/cutlass_extensions/include/cutlass_extensions/gemm/kernel/moe_problem_visitor.h:149

↓ 2 callersFunctionapply_act

tests/moe/th_moe_unit_tests.py:39

↓ 2 callersMethodapply_gradients

See base class.

examples/tensorflow/bert/bert-quantization/optimization.py:305

↓ 2 callersFunctionas_CPU_tensor

src/fastertransformer/triton_backend/triton_utils.hpp:53

↓ 2 callersMethodasdict

examples/pytorch/gpt/bloom_lambada.py:88

↓ 2 callersFunctionattention_layer

Performs multi-headed attention from `from_tensor` to `to_tensor`. This is an implementation of multi-headed attention based on "Attention is all

examples/tensorflow/bert/tensorflow_bert/my_modeling.py:567

↓ 2 callersFunctionbleu_score

(pred_file, ref_file)

examples/tensorflow/decoding/utils/bleu_score.py:19

↓ 2 callersFunctionbuild_embeddings

Args: opt: the option in current environment. text_field(TextMultiField): word and feats field. for_encoder(bool): build

examples/pytorch/decoding/utils/translation_model.py:34

↓ 2 callersFunctionbuild_hf_longformer

examples/pytorch/longformer/longformer_qa.py:51

↓ 2 callersFunctionbuild_sequence_mask

Builds the dot product mask. Args: sequence_length: The sequence length. num_heads: The number of heads. maximum_length: Optional size o

examples/tensorflow/encoder/utils/encoder.py:25

↓ 2 callersFunctionbuild_val_loader

examples/pytorch/swin/Swin-Transformer-Quantization/data.py:33

↓ 2 callersFunctioncalGptGemmTestBufSizeInByte

src/fastertransformer/utils/gemm_test/gpt_gemm_func.cc:754

↓ 2 callersFunctioncalT5GemmTestBufSizeInByte

src/fastertransformer/utils/gemm_test/t5_gemm_func.cc:766

↓ 2 callersFunctioncheckByNpz

examples/cpp/xlnet/xlnet_correctness_example.cc:148

↓ 2 callersFunctioncheckMat

examples/cpp/decoding/layernorm_test.cc:42

↓ 2 callersFunctioncheckNonZero

src/fastertransformer/kernels/layernorm_fp8_kernels_test.cc:30

↓ 2 callersFunctioncheckNonZero

examples/cpp/decoding/layernorm_test.cc:27

↓ 2 callersFunctioncheckpoint_quantization

(init_dict, ths_path='../../../build/lib/libth_transformer.so', verbose=True)

examples/pytorch/vit/checkpoint_quantization.py:26

↓ 2 callersFunctioncompareTwoTensorV2

examples/cpp/bert_fp8/bert_fp8_example.cc:38

↓ 2 callersFunctioncompute_loss

(logits, positions)

examples/tensorflow/bert/bert-quantization/run_squad.py:511

↓ 2 callersFunctioncompute_metrics

(task_name, preds, labels)

examples/pytorch/bert/bert-quantization-sparsity/run_glue.py:76

↓ 2 callersFunctionconvert_and_save_parameter

Convert and save to FT parameter Split a param into tensor_para_size if needed, and save each split param at {save_dir}/{param_name}.bin or

examples/pytorch/gpt/utils/huggingface_bloom_convert.py:216

↓ 2 callersFunctionconvert_by_vocab

Converts a sequence of [tokens|ids] using the vocab.

examples/tensorflow/bert/bert-quantization/tokenization.py:130

↓ 2 callersFunctionconvert_examples_to_features

Loads a data file into a list of `InputBatch`s.

examples/tensorflow/bert/bert-quantization/utils/create_squad_data.py:306

↓ 2 callersFunctionconvert_examples_to_features

Loads a data file into a list of `InputBatch`s.

examples/pytorch/bert/bert-quantization-sparsity/run_squad.py:226

↓ 2 callersFunctionconvert_examples_to_features

Loads a data file into a list of `InputBatch`s.

examples/pytorch/bert/bert-quantization-sparsity/processors/glue.py:230

↓ 2 callersFunctionconvert_single_example

Converts a single `InputExample` into a single `InputFeatures`.

examples/tensorflow/bert/bert-quantization/utils/create_glue_data.py:332

↓ 2 callersFunctioncopy_tiles_and_advance

src/fastertransformer/cutlass_extensions/include/cutlass_extensions/gemm/threadblock/dq_mma_multistage.h:98

↓ 2 callersFunctioncorresponding_ft_name

(ia3_weight, ia3_name=False)

examples/pytorch/t5/utils/nemo_t5_ia3.py:148

↓ 2 callersMethodcustom_moe_fc

(self, torch_activations, torch_experts_for_rows, torch_weights, torch_scales, torch_biases, activation_str)

tests/moe/th_moe_unit_tests.py:135

↓ 2 callersMethoddecode

(self, token_ids)

examples/pytorch/gpt/utils/tokenizer.py:67

↓ 2 callersFunctiondecoding_body

(word_ids, step, memory, memory_sequence_length,

examples/tensorflow/decoder/utils/decoding.py:97

↓ 2 callersMethoddequantize

src/fastertransformer/cutlass_extensions/include/cutlass_extensions/gemm/warp/mma_tensorop_dequantizer.h:259

↓ 2 callersMethoddeserialize

src/fastertransformer/tensorrt_plugin/swin/serialize.hpp:74

↓ 2 callersMethoddownload

examples/pytorch/bert/bert-quantization-sparsity/data/Downloader.py:28

↓ 2 callersMethoddownload_bookscorpus

examples/pytorch/bert/bert-quantization-sparsity/data/Downloader.py:64

↓ 2 callersMethoddownload_google_pretrained_weights

examples/pytorch/bert/bert-quantization-sparsity/data/Downloader.py:72

↓ 2 callersMethoddownload_nvidia_pretrained_weights

examples/pytorch/bert/bert-quantization-sparsity/data/Downloader.py:76

↓ 2 callersMethoddownload_squad

examples/pytorch/bert/bert-quantization-sparsity/data/Downloader.py:84

↓ 2 callersMethodencode

examples/pytorch/gpt/utils/tokenizer.py:56

↓ 2 callersFunctionextract_amaxlist

(init_dict, depths, version=1, ths_path='../../../build/lib/libth_transformer.so', verbose=True)

examples/pytorch/swin/checkpoint_quantization.py:27

↓ 2 callersFunctionfixprob

examples/tensorflow/bert/bert-quantization/run_squad.py:519

↓ 2 callersFunctionflatten_adapter

Flatten adapter weights.

examples/pytorch/t5/utils/nemo_t5_ckpt_convert.py:655

↓ 2 callersMethodforward

src/fastertransformer/models/bart/BartDecoder.cc:291

↓ 2 callersMethodforward

src/fastertransformer/models/longformer/LongformerEncoder.cc:140

↓ 2 callersMethodforward

src/fastertransformer/models/t5/T5Decoding.cc:355

↓ 2 callersMethodforward

(self, input_ids, seq_len, remove_padding=True)

examples/tensorflow/deberta/utils/ft_deberta.py:160

↓ 2 callersMethodforward

(self, *args, **kwargs)

examples/pytorch/longformer/model.py:83

↓ 2 callersMethodfrom_json_file

Constructs a `BertConfig` from a json file of parameters.

examples/tensorflow/bert/tensorflow_bert/my_modeling.py:94

↓ 2 callersFunctionftNcclCommInitRank

src/fastertransformer/utils/nccl_utils.cc:282

↓ 2 callersFunctionfuse_qkv

Adjust quantization ranges to match an implementation where the QKV projections are implemented with a single GEMM. Force the weight and output sc

examples/pytorch/vit/ViT-quantization/quant_utils.py:189

↓ 2 callersFunctionfuse_qkv

Adjust quantization ranges to match an implementation where the QKV projections are implemented with a single GEMM. Force the weight and output sc

examples/pytorch/bert/bert-quantization-sparsity/quant_utils.py:189

↓ 2 callersFunctionfuse_qkv

Adjust quantization ranges to match an implementation where the QKV projections are implemented with a single GEMM. Force the weight and output sc

examples/pytorch/swin/Swin-Transformer-Quantization/quant_utils.py:160

↓ 2 callersFunctiongather_nd

(params, indices)

examples/pytorch/decoding/utils/decoding.py:141

↓ 2 callersMethodgemm_bias_act

src/fastertransformer/kernels/cutlass_kernels/fpA_intB_gemm/fpA_intB_gemm_template.h:511

↓ 2 callersFunctiongen_tensor_dataset

examples/pytorch/bert/bert-quantization-sparsity/run_glue.py:406

↓ 2 callersFunctiongenerate_encoder_igemm_config

src/fastertransformer/utils/gemm_test/encoder_igemm_func.cc:1043

↓ 2 callersFunctiongenerate_encoder_result

(batch_size, max_seq_len, memory_hidden_dim,

examples/tensorflow/decoder/utils/decoding.py:53

↓ 2 callersMethodgetHiddenUnits

src/fastertransformer/models/multi_gpu_gpt/ParallelGpt.cc:1763

↓ 2 callersMethodgetNumHeads

src/fastertransformer/models/t5/T5Encoder.h:167

↓ 2 callersMethodgetNumIA3Tasks

src/fastertransformer/models/t5/T5EncoderWeight.h:66

↓ 2 callersMethodgetNumLayers

src/fastertransformer/models/t5/T5Encoder.h:162

↓ 2 callersMethodgetStep

src/fastertransformer/models/multi_gpu_gpt/ParallelGpt.cc:1775

↓ 2 callersFunctionget_answer_text

(example, feature, pred, args)

examples/pytorch/bert/bert-quantization-sparsity/run_squad.py:524

↓ 2 callersFunctionget_calibrators

Prepare collector and relevant tensors for calibration and return a list of calibrators. Args: collection_name_prefix: A string. Determine the

examples/tensorflow/bert/bert-quantization/ft-tensorflow-quantization/ft_tensorflow_quantization/python/calib/calibrator.py:299

↓ 2 callersFunctionget_file_name

(idx_moe, idx_expert, idx_rank)

examples/pytorch/gpt/utils/megatron_gpt_moe_ckpt_convert.py:345

↓ 2 callersMethodget_knowledge_distillation_loss

(self, output_student, output_teacher)

examples/pytorch/bert/bert-quantization-sparsity/run_squad.py:71

↓ 2 callersMethodget_labels

Gets the list of labels for this data set.

examples/tensorflow/bert/bert-quantization/utils/create_glue_data.py:131

↓ 2 callersMethodget_lr

examples/pytorch/bert/bert-quantization-sparsity/schedulers.py:126

↓ 2 callersFunctionget_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

examples/pytorch/gpt/utils/huggingface_bloom_convert.py:335

↓ 2 callersFunctionget_op_cache_config

(size_per_head, dtype)

examples/tensorflow/decoder/utils/decoder.py:21

↓ 2 callersFunctionget_pairs

Return set of symbol pairs in a word. Word is represented as tuple of symbols (symbols being variable-length strings).

examples/tensorflow/gpt/utils/gpt_token_encoder.py:68

↓ 2 callersFunctionget_pairs

Return set of symbol pairs in a word. Word is represented as tuple of symbols (symbols being variable-length strings).

examples/pytorch/gpt/utils/gpt_token_encoder.py:68

↓ 2 callersMethodget_sequence_output

Gets final hidden layer of encoder. Returns: float Tensor of shape [batch_size, seq_length, hidden_size] corresponding to the final h

examples/tensorflow/bert/tensorflow_bert/my_modeling.py:243

↓ 2 callersMethodget_sequence_output

Gets final hidden layer of encoder. Returns: float Tensor of shape [batch_size, seq_length, hidden_size] corresponding to the final h

examples/tensorflow/bert/bert-quantization/modeling.py:253

↓ 2 callersMethodget_test_examples

Gets a collection of `InputExample`s for prediction.

examples/tensorflow/bert/bert-quantization/utils/create_glue_data.py:127

↓ 2 callersMethodget_tokenizer_file_path

(self, tokenizer_key, file_key, default_filename_pattern)

examples/pytorch/nemo.py:178

↓ 2 callersFunctionget_valid_prelim_predictions

(start_indices, end_indices, feature, result, args)

examples/pytorch/bert/bert-quantization-sparsity/run_squad.py:543

↓ 2 callersFunctionget_weight_data_type

examples/pytorch/t5/utils/megatron_t5_ckpt_convert.py:95

↓ 2 callersMethodget_workspace_size

src/fastertransformer/cutlass_extensions/include/cutlass_extensions/gemm/kernel/moe_problem_visitor.h:323

↓ 2 callersFunctionhmma_fp32

src/fastertransformer/kernels/decoder_masked_multihead_attention/decoder_masked_multihead_attention_template.hpp:632

↓ 2 callersFunctionini_parse_file

See documentation in header file. */

3rdparty/INIReader.h:275

↓ 2 callersFunctioninit_graph

examples/tensorflow/ckpt_type_convert.py:28

↓ 2 callersFunctioninit_graph

examples/tensorflow/bert/tensorflow_bert/ckpt_type_convert.py:28

↓ 2 callersFunctioninit_op_cache

(decoder_args, batchxbeam, memory_max_seq_len, decoding_max_seq_len)

examples/tensorflow/decoder/utils/decoder.py:304

↓ 2 callersFunctioninitialize_decoding_variables

(decoding_args, batchxbeam)

examples/tensorflow/decoder/utils/decoding.py:30

↓ 2 callersFunctioninput_fn_builder

Creates an `input_fn` closure to be passed to Estimator.

examples/tensorflow/bert/bert-quantization/run_squad.py:571

← previous next →501–600 of 4,228, ranked by callers