Buckets:

|
download
raw
7.88 kB

CPM

Overview

CPM モデルは、Zhengyan Zhang、Xu Han、Hao Zhou、Pei Ke、Yuxian Gu によって CPM: A Large-scale Generative Chinese Pre-trained Language Model で提案されました。葉徳明、秦裕佳、 Yusheng Su、Haozhe Ji、Jian Guan、Fanchao Qi、Xiaozi Wang、Yanan Zheng、Guoyang Zeng、Huanqi Cao、Shengqi Chen、 Daixuan Li、Zhenbo Sun、Zhiyuan Liu、Minlie Huang、Wentao Han、Jie Tang、Juanzi Li、Xiaoyan Zhu、Maosong Sun。

論文の要約は次のとおりです。

事前トレーニングされた言語モデル (PLM) は、さまざまな下流の NLP タスクに有益であることが証明されています。最近ではGPT-3、 1,750億個のパラメータと570GBの学習データを備え、数回の撮影(1枚でも)の容量で大きな注目を集めました ゼロショット)学習。ただし、GPT-3 を適用して中国語の NLP タスクに対処することは依然として困難です。 GPT-3 の言語は主に英語であり、パラメーターは公開されていません。この技術レポートでは、 大規模な中国語トレーニング データに対する生成的事前トレーニングを備えた中国語事前トレーニング済み言語モデル (CPM)。最高に 私たちの知識の限りでは、26 億のパラメータと 100GB の中国語トレーニング データを備えた CPM は、事前トレーニングされた中国語としては最大のものです。 言語モデルは、会話、エッセイの作成、 クローゼテストと言語理解。広範な実験により、CPM が多くの環境で優れたパフォーマンスを達成できることが実証されています。 少数ショット (ゼロショットでも) 学習の設定での NLP タスク。

このモデルは canwenxu によって提供されました。オリジナルの実装が見つかります ここ: https://github.com/TsinghuaAI/CPM-Generate

CPM のアーキテクチャは、トークン化方法を除いて GPT-2 と同じです。詳細については、GPT-2 ドキュメント を参照してください。 API リファレンス情報。

CpmTokenizer[[transformers.CpmTokenizer]]

transformers.CpmTokenizer[[transformers.CpmTokenizer]]

Source

Runs pre-tokenization with Jieba-RS segmentation tool. It is used in CPM models.

build_inputs_with_special_tokenstransformers.CpmTokenizer.build_inputs_with_special_tokenshttps://github.com/huggingface/transformers/blob/main/src/transformers/models/cpm/tokenization_cpm.py#L230[{"name": "token_ids_0", "val": ": list"}, {"name": "token_ids_1", "val": ": list[int] | None = None"}]- token_ids_0 (list[int]) -- List of IDs to which the special tokens will be added.

  • token_ids_1 (list[int], optional) -- Optional second list of IDs for sequence pairs.0list[int]List of input IDs with the appropriate special tokens.

Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and adding special tokens. An XLNet sequence has the following format:

  • single sequence: X
  • pair of sequences: A B

Parameters:

token_ids_0 (list[int]) : List of IDs to which the special tokens will be added.

token_ids_1 (list[int], optional) : Optional second list of IDs for sequence pairs.

Returns:

list[int]

List of input IDs with the appropriate special tokens.

convert_tokens_to_string[[transformers.CpmTokenizer.convert_tokens_to_string]]

Source

Converts a sequence of tokens (strings for sub-words) in a single string.

create_token_type_ids_from_sequences[[transformers.CpmTokenizer.create_token_type_ids_from_sequences]]

Source

Create a mask from the two sequences passed to be used in a sequence-pair classification task. An XLNet

sequence pair mask has the following format:

0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1
| first sequence    | second sequence |

If token_ids_1 is None, this method only returns the first portion of the mask (0s).

Parameters:

token_ids_0 (list[int]) : List of IDs.

token_ids_1 (list[int], optional) : Optional second list of IDs for sequence pairs.

Returns:

list[int]

List of token type IDs according to the given sequence(s).

get_special_tokens_mask[[transformers.CpmTokenizer.get_special_tokens_mask]]

Source

Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding special tokens using the tokenizer prepare_for_model method.

Parameters:

token_ids_0 (list[int]) : List of IDs.

token_ids_1 (list[int], optional) : Optional second list of IDs for sequence pairs.

already_has_special_tokens (bool, optional, defaults to False) : Whether or not the token list is already formatted with special tokens for the model.

Returns:

list[int]

A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.

CpmTokenizerFast[[transformers.CpmTokenizerFast]]

transformers.CpmTokenizerFast[[transformers.CpmTokenizerFast]]

Source

Runs pre-tokenization with Jieba-RS segmentation tool. It is used in CPM models.

build_inputs_with_special_tokenstransformers.CpmTokenizerFast.build_inputs_with_special_tokenshttps://github.com/huggingface/transformers/blob/main/src/transformers/models/cpm/tokenization_cpm_fast.py#L145[{"name": "token_ids_0", "val": ": list"}, {"name": "token_ids_1", "val": ": list[int] | None = None"}]- token_ids_0 (list[int]) -- List of IDs to which the special tokens will be added.

  • token_ids_1 (list[int], optional) -- Optional second list of IDs for sequence pairs.0list[int]List of input IDs with the appropriate special tokens.

Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and adding special tokens. An XLNet sequence has the following format:

  • single sequence: X
  • pair of sequences: A B

Parameters:

token_ids_0 (list[int]) : List of IDs to which the special tokens will be added.

token_ids_1 (list[int], optional) : Optional second list of IDs for sequence pairs.

Returns:

list[int]

List of input IDs with the appropriate special tokens.

create_token_type_ids_from_sequences[[transformers.CpmTokenizerFast.create_token_type_ids_from_sequences]]

Source

Create a mask from the two sequences passed to be used in a sequence-pair classification task. An XLNet

sequence pair mask has the following format:

0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1
| first sequence    | second sequence |

If token_ids_1 is None, this method only returns the first portion of the mask (0s).

Parameters:

token_ids_0 (list[int]) : List of IDs.

token_ids_1 (list[int], optional) : Optional second list of IDs for sequence pairs.

Returns:

list[int]

List of token type IDs according to the given sequence(s).

Xet Storage Details

Size:
7.88 kB
·
Xet hash:
17007a7f6c58ca1712054ff19697670274b94bb79c454e11bc8f7a730486b4e2

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.