Functions — github.com/OpenBMB/CPM-Bee

↓ 1 callersFunctionget_args

()

src/build_dataset.py:13

↓ 1 callersFunctionget_learning_rate_scheduler

(args, optimizer)

src/pretrain_cpm_bee.py:59

↓ 1 callersFunctionget_learning_rate_scheduler

(args, optimizer)

src/finetune_cpm_bee.py:60

↓ 1 callersFunctionget_model

(args)

src/pretrain_cpm_bee.py:35

↓ 1 callersFunctionget_model

(args)

src/finetune_cpm_bee.py:35

↓ 1 callersFunctionget_optimizer

(args, model)

src/pretrain_cpm_bee.py:45

↓ 1 callersFunctionget_optimizer

(args, model)

src/finetune_cpm_bee.py:53

↓ 1 callersMethodget_piece

(self, text: str)

src/cpm_live/tokenizers/bee.py:75

↓ 1 callersFunctionget_tokenizer

(args)

src/pretrain_cpm_bee.py:30

↓ 1 callersFunctionget_tokenizer

(args)

src/finetune_cpm_bee.py:30

↓ 1 callersFunctioninitialize

()

src/pretrain_cpm_bee.py:88

↓ 1 callersFunctioninitialize

()

src/finetune_cpm_bee.py:89

↓ 1 callersFunctionload_dataset_config

(dataset_path: str)

src/cpm_live/utils/config.py:23

↓ 1 callersFunctionload_delta

(delta_path)

src/text_generation_hf.py:16

↓ 1 callersFunctionload_vocab

Loads a vocabulary file into a dictionary.

src/cpm_live/tokenizers/ant.py:22

↓ 1 callersFunctionload_vocab

Loads a vocabulary file into a dictionary.

src/cpm_live/tokenizers/bee.py:21

↓ 1 callersFunctionmain

()

src/pretrain_cpm_bee.py:382

↓ 1 callersFunctionmain

()

src/preprocess_dataset.py:24

↓ 1 callersFunctionmain

()

src/build_dataset.py:56

↓ 1 callersFunctionmain

()

src/text_generation_hf.py:28

↓ 1 callersFunctionmain

()

src/finetune_cpm_bee.py:424

↓ 1 callersFunctionmain

()

src/text_generation.py:17

↓ 1 callersFunctionparse_args

()

src/text_generation_hf.py:8

↓ 1 callersFunctionparse_args

()

src/text_generation.py:8

↓ 1 callersFunctionpre_process

(line)

src/build_dataset.py:47

↓ 1 callersFunctionpretrain

( args, tokenizer: CPMBeeTokenizer, model: CPMBee, optimizer: bmt.optim.AdamOffloadOptimizer,

src/pretrain_cpm_bee.py:155

↓ 1 callersFunctionreformat_data

set your data format

src/preprocess_dataset.py:19

↓ 1 callersFunctionrename_if_exists

(file_path)

src/cpm_live/utils/export.py:12

↓ 1 callersFunctionrms_layernorm

(hidden: torch.Tensor, weight: torch.Tensor, eps: float)

src/cpm_live/layers/layernorm.py:6

↓ 1 callersFunctionrms_layernorm

(hidden: torch.Tensor, weight: torch.Tensor, eps: float)

src/cpm_live/native_layers/layernorm.py:5

↓ 1 callersFunctionsee_memory

(detail=False)

src/pretrain_cpm_bee.py:97

↓ 1 callersFunctionsee_memory

(detail=False)

src/finetune_cpm_bee.py:97

↓ 1 callersMethodserialize

(self, obj)

src/cpm_live/dataset/serializer.py:24

↓ 1 callersFunctionsetup_model_and_optimizer

(args)

src/pretrain_cpm_bee.py:72

↓ 1 callersFunctionsetup_model_and_optimizer

(args)

src/finetune_cpm_bee.py:73

↓ 1 callersFunctionshuffle_dataset

Shuffle one distributed datataset, write results to another dataset. Args: path_str (str): path to source dataset path_tgt (str):

src/cpm_live/dataset/utils.py:43

↓ 1 callersFunctionsplit_sent

(data_, depth, args, seg)

src/build_dataset.py:26

↓ 1 callersMethodstate_dict

(self)

src/cpm_live/training_tasks/bee/pretrain.py:986

↓ 1 callersMethodto_dict

(self)

src/cpm_live/utils/config.py:75

↓ 1 callersMethodto_json_string

(self)

src/cpm_live/utils/config.py:71

↓ 1 callersMethodtokenize

Tokenize a string.

src/cpm_live/tokenizers/ant.py:135

↓ 1 callersMethodtokenize

(self, text: str)

src/cpm_live/tokenizers/bee.py:107

↓ 1 callersFunctiontop_k_top_p_filtering

(logits, top_k=0, top_p=0.0, filter_value=-float("inf"))

src/cpm_live/generation/generation_utils.py:5

↓ 1 callersMethodupdate_data

(self, data: Any)

src/pretrain_cpm_bee.py:122

↓ 1 callersMethodupdate_loss

(self, iteration: int, loss_map: Dict[str, float])

src/pretrain_cpm_bee.py:127

Method__del__

(self)

src/cpm_live/training_tasks/bee/pretrain.py:1036

Method__enter__

(self)

src/cpm_live/dataset/distributed_dataset.py:711

Method__exit__

(self, exc_type, exc_value, exc_traceback)

src/cpm_live/dataset/distributed_dataset.py:715

Method__init__

(self, log_path: str)

src/pretrain_cpm_bee.py:117

Method__init__

(self, vocab, unk_token="<unk>", max_input_chars_per_word=200)

src/cpm_live/tokenizers/ant.py:36

Method__init__

( self, bod_token="<d>", eod_token="</d>", bos_token="<s>", eos_token=

src/cpm_live/tokenizers/ant.py:68

Method__init__

(self, token: str, start: int, is_unk: bool, is_special: bool)

src/cpm_live/tokenizers/bee.py:36

Method__init__

( self, )

src/cpm_live/tokenizers/bee.py:52

Method__init__

(self)

src/cpm_live/utils/config.py:50

Method__init__

(self, path: str)

src/cpm_live/utils/log.py:31

Method__init__

(self, fname, mode)

src/cpm_live/dataset/distributed_dataset.py:137

Method__init__

( self, path: str, rank: int = 0, world_size: int = 1, serializer: Opt

src/cpm_live/dataset/distributed_dataset.py:223

Method__init__

( self, path: str, serializer: Optional[Serializer] = None, shuffle: bool = Tr

src/cpm_live/dataset/distributed_dataset.py:589

Method__init__

(self, fname: str, block_size: int, serializer: Optional[Serializer] = None)

src/cpm_live/dataset/distributed_dataset.py:618

Method__init__

( self, path: str, dbname: str, block_size=_DEFAULT_BLOCK_SIZE, serial

src/cpm_live/dataset/distributed_dataset.py:680

Method__init__

(self)

src/cpm_live/dataset/serializer.py:21

Method__init__

(self)

src/cpm_live/dataset/serializer.py:32

Method__init__

(self)

src/cpm_live/dataset/serializer.py:43

Method__init__

(self)

src/cpm_live/dataset/serializer.py:54

Method__init__

(self, model, tokenizer, prompt_length=32)

src/cpm_live/generation/ant.py:8

Method__init__

Initialize n-best list of hypotheses.

src/cpm_live/generation/generation_utils.py:70

Method__init__

(self, model: CPMBee, tokenizer: CPMBeeTokenizer)

src/cpm_live/generation/bee.py:13

Method__init__

( self, dataset_path: str, batch_size: int, max_length: int, tokenizer

src/cpm_live/training_tasks/bee/finetune.py:8

Method__init__

( self, batch_size: int, max_length: int, tokenizer: CPMBeeTokenizer,

src/cpm_live/training_tasks/bee/pretrain.py:253

Method__init__

(self, config_path: str)

src/cpm_live/training_tasks/bee/pretrain.py:701

Method__init__

( self, config_path: str, batch_size: int, max_length: int, tokenizer:

src/cpm_live/training_tasks/bee/pretrain.py:944

Method__init__

(self, ctx, max_length=1024, prompt_length=32, tokenizer=None)

src/cpm_live/training_tasks/ant/pretrain.py:22

Method__init__

( self, dim_norm: int, dtype: torch.dtype = torch.half, eps: float = 1e-6,

src/cpm_live/layers/layernorm.py:16

Method__init__

( self, dim_model: int, num_heads: int, dim_head: int, dtype=torch.hal

src/cpm_live/layers/blocks.py:36

Method__init__

( self, dim_model: int, dim_ff: int, num_heads: int, dim_head: int,

src/cpm_live/layers/blocks.py:174

Method__init__

( self, dim_in: int, dim_out: int, dtype: torch.dtype = torch.half, in

src/cpm_live/layers/linear.py:23

Method__init__

( self, num_heads: int, num_segments: int = 1, num_buckets: int = 32,

src/cpm_live/layers/position_embedding.py:23

Method__init__

( self, num_heads: int, num_buckets: int = 32, num_segment_bucket: int = 32,

src/cpm_live/layers/position_embedding.py:132

Method__init__

( self, dim_model: int, num_heads: int, dim_head: int, dtype: torch.dt

src/cpm_live/layers/attention.py:24

Method__init__

( self, num_layers: int, dim_model: int, dim_ff: int, num_heads: int,

src/cpm_live/layers/transformer.py:38

Method__init__

( self, vocab_size: int, embedding_size: int, dtype: torch.dtype = torch.half,

src/cpm_live/layers/embedding.py:68

Method__init__

( self, dim_in: int, dim_ff: int, dtype=torch.half, )

src/cpm_live/layers/feedforward.py:23

Method__init__

( self, vocab_size=30720, dim_model=4096, num_heads=64, dim_head=64,

src/cpm_live/models/ant.py:24

Method__init__

(self, config: CPMBeeConfig)

src/cpm_live/models/bee_torch.py:24

Method__init__

(self, config: CPMAntConfig)

src/cpm_live/models/ant_torch.py:21

Method__init__

( self, vocab_size=30720, dim_model=4096, num_heads=64, dim_head=64,

src/cpm_live/models/bee.py:36

Method__init__

( self, dim_norm: int, dtype: torch.dtype = torch.half, eps: float = 1e-6,

src/cpm_live/native_layers/layernorm.py:15

Method__init__

( self, dim_model: int, num_heads: int, dim_head: int, dtype=torch.hal

src/cpm_live/native_layers/blocks.py:35

Method__init__

( self, dim_model: int, dim_ff: int, num_heads: int, dim_head: int,

src/cpm_live/native_layers/blocks.py:173

Method__init__

( self, dim_in: int, dim_out: int, dtype: torch.dtype = torch.half, in

src/cpm_live/native_layers/linear.py:22

Method__init__

( self, num_heads, num_segments=1, num_buckets=32, max_distance=128,

src/cpm_live/native_layers/position_embedding.py:22

Method__init__

( self, num_heads: int, num_buckets: int = 32, num_segment_bucket: int = 32,

src/cpm_live/native_layers/position_embedding.py:128

Method__init__

( self, dim_model: int, num_heads: int, dim_head: int, dtype: torch.dt

src/cpm_live/native_layers/attention.py:23

Method__init__

( self, num_layers: int, dim_model: int, dim_ff: int, num_heads: int,

src/cpm_live/native_layers/transformer.py:37

Method__init__

( self, vocab_size: int, embedding_size: int, dtype: torch.dtype = torch.half,

src/cpm_live/native_layers/embedding.py:64

Method__init__

( self, dim_in: int, dim_ff: int, dtype=torch.half, )

src/cpm_live/native_layers/feedforward.py:22

Method__iter__

(self)

src/cpm_live/dataset/distributed_dataset.py:604

Method__iter__

(self)

src/cpm_live/training_tasks/bee/finetune.py:54

Method__iter__

(self)

src/cpm_live/training_tasks/bee/pretrain.py:1032

Method__iter__

(self)

src/cpm_live/training_tasks/ant/pretrain.py:105

Functions302 in github.com/OpenBMB/CPM-Bee