From 9529f89abd41ee7ef0d9e2e32596ef9ee1aedb1e Mon Sep 17 00:00:00 2001 From: yh Date: Fri, 30 Aug 2019 19:54:28 +0800 Subject: [PATCH] =?UTF-8?q?=E5=A2=9E=E5=8A=A0DataBundle=E7=9A=84=E6=96=B9?= =?UTF-8?q?=E6=B3=95=EF=BC=9B=E5=A2=9E=E5=8A=A0BilSTMCRF=E7=9A=84=E6=B3=A8?= =?UTF-8?q?=E9=87=8A?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- fastNLP/core/dataset.py | 14 ++-- fastNLP/io/data_bundle.py | 72 +++++++++++++++- fastNLP/models/sequence_labeling.py | 83 ++++++++----------- .../seqence_labelling/ner/train_ontonote.py | 4 +- 4 files changed, 112 insertions(+), 61 deletions(-) diff --git a/fastNLP/core/dataset.py b/fastNLP/core/dataset.py index 51bcef43..551cf1f8 100644 --- a/fastNLP/core/dataset.py +++ b/fastNLP/core/dataset.py @@ -575,18 +575,18 @@ class DataSet(object): """ return len(self) - def rename_field(self, old_name, new_name): + def rename_field(self, field_name, new_field_name): """ 将某个field重新命名. - :param str old_name: 原来的field名称。 - :param str new_name: 修改为new_name。 + :param str field_name: 原来的field名称。 + :param str new_field_name: 修改为new_name。 """ - if old_name in self.field_arrays: - self.field_arrays[new_name] = self.field_arrays.pop(old_name) - self.field_arrays[new_name].name = new_name + if field_name in self.field_arrays: + self.field_arrays[new_field_name] = self.field_arrays.pop(field_name) + self.field_arrays[new_field_name].name = new_field_name else: - raise KeyError("DataSet has no field named {}.".format(old_name)) + raise KeyError("DataSet has no field named {}.".format(field_name)) return self def set_target(self, *field_names, flag=True, use_1st_ins_infer_dim_type=True): diff --git a/fastNLP/io/data_bundle.py b/fastNLP/io/data_bundle.py index 969730a3..f30add34 100644 --- a/fastNLP/io/data_bundle.py +++ b/fastNLP/io/data_bundle.py @@ -139,9 +139,44 @@ class DataBundle: dataset.set_target(field_name, flag=flag, use_1st_ins_infer_dim_type=use_1st_ins_infer_dim_type) return self + def set_pad_val(self, field_name, pad_val, ignore_miss_dataset=True): + """ + 将DataBundle中所有的DataSet中名为field_name的Field的padding值设置为pad_val. + + :param str field_name: + :param int pad_val: + :param bool ignore_miss_dataset: 当某个field名称在某个dataset不存在时，如果为True，则直接忽略该DataSet; + 如果为False，则报错 + :return: self + """ + for name, dataset in self.datasets.items(): + if dataset.has_field(field_name=field_name): + dataset.set_pad_val(field_name=field_name, pad_val=pad_val) + elif not ignore_miss_dataset: + raise KeyError(f"{field_name} not found DataSet:{name}.") + return self + + def set_ignore_type(self, *field_names, flag=True, ignore_miss_dataset=True): + """ + 将DataBundle中所有的DataSet中名为*field_names的Field的ignore_type设置为flag状态 + + :param str field_names: + :param bool flag: + :param bool ignore_miss_dataset: 当某个field名称在某个dataset不存在时，如果为True，则直接忽略该DataSet; + 如果为False，则报错 + :return: self + """ + for name, dataset in self.datasets.items(): + for field_name in field_names: + if dataset.has_field(field_name=field_name): + dataset.set_ignore_type(field_name, flag=flag) + elif not ignore_miss_dataset: + raise KeyError(f"{field_name} not found DataSet:{name}.") + return self + def copy_field(self, field_name, new_field_name, ignore_miss_dataset=True): """ - 将DataBundle中所有的field_name复制一份叫new_field_name. + 将DataBundle中所有的DataSet中名为field_name的Field复制一份并命名为叫new_field_name. :param str field_name: :param str new_field_name: @@ -156,9 +191,42 @@ class DataBundle: raise KeyError(f"{field_name} not found DataSet:{name}.") return self + def rename_field(self, field_name, new_field_name, ignore_miss_dataset=True): + """ + 将DataBundle中所有DataSet中名为field_name的field重命名为new_field_name. + + :param str field_name: + :param str new_field_name: + :param bool ignore_miss_dataset: 当某个field名称在某个dataset不存在时，如果为True，则直接忽略该DataSet; + 如果为False，则报错 + :return: self + """ + for name, dataset in self.datasets.items(): + if dataset.has_field(field_name=field_name): + dataset.rename_field(field_name=field_name, new_field_name=new_field_name) + elif not ignore_miss_dataset: + raise KeyError(f"{field_name} not found DataSet:{name}.") + return self + + def delete_field(self, field_name, ignore_miss_dataset=True): + """ + 将DataBundle中所有DataSet中名为field_name的field删除掉. + + :param str field_name: + :param bool ignore_miss_dataset: 当某个field名称在某个dataset不存在时，如果为True，则直接忽略该DataSet; + 如果为False，则报错 + :return: self + """ + for name, dataset in self.datasets.items(): + if dataset.has_field(field_name=field_name): + dataset.delete_field(field_name=field_name) + elif not ignore_miss_dataset: + raise KeyError(f"{field_name} not found DataSet:{name}.") + return self + def apply_field(self, func, field_name:str, new_field_name:str, ignore_miss_dataset=True, **kwargs): """ - 对DataBundle中所有的dataset使用apply方法 + 对DataBundle中所有的dataset使用apply_field方法 :param callable func: input是instance中名为 `field_name` 的field的内容。 :param str field_name: 传入func的是哪个field。 diff --git a/fastNLP/models/sequence_labeling.py b/fastNLP/models/sequence_labeling.py index 0dff21f0..0c573a90 100644 --- a/fastNLP/models/sequence_labeling.py +++ b/fastNLP/models/sequence_labeling.py @@ -4,7 +4,7 @@ __all__ = [ "SeqLabeling", "AdvSeqLabel", - # "BiLSTMCRF" + "BiLSTMCRF" ] import torch @@ -14,7 +14,6 @@ import torch.nn.functional as F from .base_model import BaseModel from ..core.const import Const as C from ..core.utils import seq_len_to_mask -from ..embeddings import embedding from ..embeddings import get_embeddings from ..modules import ConditionalRandomField from ..modules import LSTM @@ -24,18 +23,15 @@ from ..modules.decoder.crf import allowed_transitions class BiLSTMCRF(BaseModel): """ - 结构为BiLSTM + FC + Dropout + CRF. + 结构为embedding + BiLSTM + FC + Dropout + CRF. - .. todo:: - 继续补充文档 - - :param embed: tuple: - :param num_classes: - :param num_layers: - :param hidden_size: - :param dropout: - :param target_vocab: - :param encoding_type: + :param embed: 支持(1)fastNLP的各种Embedding, (2) tuple, 指明num_embedding, dimension, 如(1000, 100) + :param num_classes: 一共多少个类 + :param num_layers: BiLSTM的层数 + :param hidden_size: BiLSTM的hidden_size，实际hidden size为该值的两倍(前向、后向) + :param dropout: dropout的概率，0为不dropout + :param target_vocab: Vocabulary对象，target与index的对应关系 + :param encoding_type: encoding的类型，支持'bioes', 'bmes', 'bio', 'bmeso'等 """ def __init__(self, embed, num_classes, num_layers=1, hidden_size=100, dropout=0.5, target_vocab=None, encoding_type=None): @@ -86,21 +82,20 @@ class SeqLabeling(BaseModel): 一个基础的Sequence labeling的模型。 用于做sequence labeling的基础类。结构包含一层Embedding，一层LSTM(单向，一层)，一层FC，以及一层CRF。 - :param tuple(int,int),torch.FloatTensor,nn.Embedding,numpy.ndarray init_embed: Embedding的大小(传入tuple(int, int), - 第一个int为vocab_zie, 第二个int为embed_dim); 如果为Tensor, Embedding, ndarray等则直接使用该值初始化Embedding + :param tuple(int,int),torch.FloatTensor,nn.Embedding,numpy.ndarray embed: Embedding的大小(传入tuple(int, int), + 第一个int为vocab_zie, 第二个int为embed_dim); 如果为Tensor, embedding, ndarray等则直接使用该值初始化Embedding :param int hidden_size: LSTM隐藏层的大小 :param int num_classes: 一共有多少类 """ - def __init__(self, init_embed, hidden_size, num_classes): + def __init__(self, embed, hidden_size, num_classes): super(SeqLabeling, self).__init__() - self.Embedding = embedding.Embedding(init_embed) - self.Rnn = encoder.LSTM(self.Embedding.embedding_dim, hidden_size) - self.Linear = nn.Linear(hidden_size, num_classes) - self.Crf = decoder.ConditionalRandomField(num_classes) - self.mask = None - + self.embedding = get_embeddings(embed) + self.rnn = encoder.LSTM(self.embedding.embedding_dim, hidden_size) + self.fc = nn.Linear(hidden_size, num_classes) + self.crf = decoder.ConditionalRandomField(num_classes) + def forward(self, words, seq_len, target): """ :param torch.LongTensor words: [batch_size, max_len]，序列的index @@ -109,17 +104,14 @@ class SeqLabeling(BaseModel): :return y: If truth is None, return list of [decode path(list)]. Used in testing and predicting. If truth is not None, return loss, a scalar. Used in training. """ - assert words.shape[0] == seq_len.shape[0] - assert target.shape == words.shape - self.mask = self._make_mask(words, seq_len) - - x = self.Embedding(words) + mask = seq_len_to_mask(seq_len, max_len=words.size(1)) + x = self.embedding(words) # [batch_size, max_len, word_emb_dim] - x, _ = self.Rnn(x, seq_len) + x, _ = self.rnn(x, seq_len) # [batch_size, max_len, hidden_size * direction] - x = self.Linear(x) + x = self.fc(x) # [batch_size, max_len, num_classes] - return {C.LOSS: self._internal_loss(x, target)} + return {C.LOSS: self._internal_loss(x, target, mask)} def predict(self, words, seq_len): """ @@ -129,18 +121,18 @@ class SeqLabeling(BaseModel): :param torch.LongTensor seq_len: [batch_size,] :return: {'pred': xx}, [batch_size, max_len] """ - self.mask = self._make_mask(words, seq_len) + mask = seq_len_to_mask(seq_len, max_len=words.size(1)) - x = self.Embedding(words) + x = self.embedding(words) # [batch_size, max_len, word_emb_dim] - x, _ = self.Rnn(x, seq_len) + x, _ = self.rnn(x, seq_len) # [batch_size, max_len, hidden_size * direction] - x = self.Linear(x) + x = self.fc(x) # [batch_size, max_len, num_classes] - pred = self._decode(x) + pred = self._decode(x, mask) return {C.OUTPUT: pred} - def _internal_loss(self, x, y): + def _internal_loss(self, x, y, mask): """ Negative log likelihood loss. :param x: Tensor, [batch_size, max_len, tag_size] @@ -152,22 +144,15 @@ class SeqLabeling(BaseModel): y = y.long() assert x.shape[:2] == y.shape assert y.shape == self.mask.shape - total_loss = self.Crf(x, y, self.mask) + total_loss = self.crf(x, y, mask) return torch.mean(total_loss) - def _make_mask(self, x, seq_len): - batch_size, max_len = x.size(0), x.size(1) - mask = seq_len_to_mask(seq_len) - mask = mask.view(batch_size, max_len) - mask = mask.to(x).float() - return mask - - def _decode(self, x): + def _decode(self, x, mask): """ :param torch.FloatTensor x: [batch_size, max_len, tag_size] :return prediction: [batch_size, max_len] """ - tag_seq, _ = self.Crf.viterbi_decode(x, self.mask) + tag_seq, _ = self.crf.viterbi_decode(x, mask) return tag_seq @@ -177,7 +162,7 @@ class AdvSeqLabel(nn.Module): 更复杂的Sequence Labelling模型。结构为Embedding, LayerNorm, 双向LSTM(两层)，FC，LayerNorm，DropOut，FC，CRF。 - :param tuple(int,int),torch.FloatTensor,nn.Embedding,numpy.ndarray init_embed: Embedding的大小(传入tuple(int, int), + :param tuple(int,int),torch.FloatTensor,nn.Embedding,numpy.ndarray embed: Embedding的大小(传入tuple(int, int), 第一个int为vocab_zie, 第二个int为embed_dim); 如果为Tensor, Embedding, ndarray等则直接使用该值初始化Embedding :param int hidden_size: LSTM的隐层大小 :param int num_classes: 有多少个类 @@ -188,11 +173,11 @@ class AdvSeqLabel(nn.Module): :param str encoding_type: 支持"BIO", "BMES", "BEMSO", 只有在id2words不为None的情况有用。 """ - def __init__(self, init_embed, hidden_size, num_classes, dropout=0.3, id2words=None, encoding_type='bmes'): + def __init__(self, embed, hidden_size, num_classes, dropout=0.3, id2words=None, encoding_type='bmes'): super().__init__() - self.Embedding = embedding.Embedding(init_embed) + self.Embedding = get_embeddings(embed) self.norm1 = torch.nn.LayerNorm(self.Embedding.embedding_dim) self.Rnn = encoder.LSTM(input_size=self.Embedding.embedding_dim, hidden_size=hidden_size, num_layers=2, dropout=dropout, diff --git a/reproduction/seqence_labelling/ner/train_ontonote.py b/reproduction/seqence_labelling/ner/train_ontonote.py index ee80b6f7..9fd13100 100644 --- a/reproduction/seqence_labelling/ner/train_ontonote.py +++ b/reproduction/seqence_labelling/ner/train_ontonote.py @@ -18,11 +18,9 @@ from fastNLP.io.pipe.conll import OntoNotesNERPipe #######hyper normalize = False -lower = False lr = 0.01 dropout = 0.5 batch_size = 32 -job_embed = False data_name = 'ontonote' #######hyper @@ -41,7 +39,7 @@ def cache(): word_dropout=0.01, dropout=dropout, lower=True, - min_freq=2) + min_freq=1) return data, char_embed, word_embed data, char_embed, word_embed = cache()