def __init__(self, sents_src, sents_tgt, vocab_path):
# 一般init函数是加载所有数据
super(BertDataset, self).__init__()
self.sents_src = sents_src
self.sents_tgt = sents_tgt
self.word2idx = load_chinese_base_vocab(vocab_path)
self.idx2word = {k: v for v, k in self.word2idx.items()}
self.tokenizer = Tokenizer(self.word2idx)
def read_corpus(dir_path, vocab_path):
sents_src = []
sents_tgt = []
word2idx = load_chinese_base_vocab(vocab_path)
tokenizer = Tokenizer(word2idx)
files = os.listdir(dir_path)
for file1 in files:
if not os.path.isdir(file1):
file_path = dir_path + "/" + file1
print(file_path)
if file_path[-3:] != "csv":
continue
df = pd.read_csv(file_path)
for index, row in df.iterrows():
if type(row[0]) is not str or type(row[3]) is not str:
continue
if len(row[0]) > 8 or len(row[0]) < 2:
# 过滤掉题目长度过长和过短的诗句
continue
if len(row[0].split(" ")) > 1:
# 说明题目里面存在空格，只要空格前面的数据
row[0] = row[0].split(" ")[0]
encode_text = tokenizer.encode(row[3])[0]
if word2idx["[UNK]"] in encode_text:
continue
if len(row[3]) == 24 and (row[3][5] == "，"
or row[3][5] == "。"):
# 五言绝句
sents_src.append(row[0] + "##" + "五言绝句")
sents_tgt.append(row[3])
elif len(row[3]) == 32 and (row[3][7] == "，"
or row[3][7] == "。"):
# 七言绝句
sents_src.append(row[0] + "##" + "七言绝句")
sents_tgt.append(row[3])
elif len(row[3]) == 48 and (row[3][5] == "，"
or row[3][5] == "。"):
# 五言律诗
sents_src.append(row[0] + "##" + "五言律诗")
sents_tgt.append(row[3])
elif len(row[3]) == 64 and (row[3][7] == "，"
or row[3][7] == "。"):
# 七言律诗
sents_src.append(row[0] + "##" + "七言律诗")
sents_tgt.append(row[3])
print("诗句共: " + str(len(sents_src)) + "篇")
return sents_src, sents_tgt
class BertDataset(Dataset):
"""
针对特定数据集，定义一个相关的取数据的方式
"""
def __init__(self, sents_src, sents_tgt, vocab_path):
# 一般init函数是加载所有数据
super(BertDataset, self).__init__()
self.sents_src = sents_src
self.sents_tgt = sents_tgt
self.word2idx = load_chinese_base_vocab(vocab_path)
self.idx2word = {k: v for v, k in self.word2idx.items()}
self.tokenizer = Tokenizer(self.word2idx)
def __getitem__(self, i):
src = self.sents_src[i]
tgt = self.sents_tgt[i]
token_ids, token_type_ids = self.tokenizer.encode(src, tgt)
output = {
"token_ids": token_ids,
"token_type_ids": token_type_ids,
}
return output
def __len__(self):
return len(self.sents_src)
def __init__(self, vocab_path, model_name="roberta"):
super(Seq2SeqModel, self).__init__()
self.word2ix = load_chinese_base_vocab(vocab_path)
self.tokenizer = Tokenizer(self.word2ix)
config = ""
if model_name == "roberta":
from bert.seq2seq.model.roberta_model import BertModel, BertConfig, BertLMPredictionHead
config = BertConfig(len(self.word2ix))
self.bert = BertModel(config)
self.decoder = BertLMPredictionHead(
config, self.bert.embeddings.word_embeddings.weight)
elif model_name == "bert":
from bert.seq2seq.model.bert_model import BertConfig, BertModel, BertLMPredictionHead
config = BertConfig(len(self.word2ix))
self.bert = BertModel(config)
self.decoder = BertLMPredictionHead(
config, self.bert.embeddings.word_embeddings.weight)
else:
raise Exception("model_name_err")
self.hidden_dim = config.hidden_size
self.vocab_size = config.vocab_size
def __init__(self, vocab_path, target_size, model_name="roberta"):
super(BertClsClassifier, self).__init__()
self.word2ix = load_chinese_base_vocab(vocab_path)
self.tokenizer = Tokenizer(self.word2ix)
self.target_size = target_size
config = ""
if model_name == "roberta":
from bert.seq2seq.model.roberta_model import BertModel, BertConfig
config = BertConfig(len(self.word2ix))
self.bert = BertModel(config)
elif model_name == "bert":
from bert.seq2seq.model.bert_model import BertConfig, BertModel
config = BertConfig(len(self.word2ix))
self.bert = BertModel(config)
else :
raise Exception("model_name_err")
self.final_dense = nn.Linear(config.hidden_size, self.target_size)
class Seq2SeqModel(nn.Module):
"""
"""
def __init__(self, vocab_path, model_name="roberta"):
super(Seq2SeqModel, self).__init__()
self.word2ix = load_chinese_base_vocab(vocab_path)
self.tokenizer = Tokenizer(self.word2ix)
config = ""
if model_name == "roberta":
from bert.seq2seq.model.roberta_model import BertModel, BertConfig, BertLMPredictionHead
config = BertConfig(len(self.word2ix))
self.bert = BertModel(config)
self.decoder = BertLMPredictionHead(
config, self.bert.embeddings.word_embeddings.weight)
elif model_name == "bert":
from bert.seq2seq.model.bert_model import BertConfig, BertModel, BertLMPredictionHead
config = BertConfig(len(self.word2ix))
self.bert = BertModel(config)
self.decoder = BertLMPredictionHead(
config, self.bert.embeddings.word_embeddings.weight)
else:
raise Exception("model_name_err")
self.hidden_dim = config.hidden_size
self.vocab_size = config.vocab_size
def compute_loss(self, predictions, labels, target_mask):
"""
target_mask : 句子a部分和pad部分全为0， 而句子b部分为1
"""
predictions = predictions.view(-1, self.vocab_size)
labels = labels.view(-1)
target_mask = target_mask.view(-1).float()
loss = nn.CrossEntropyLoss(ignore_index=0, reduction="none")
# 通过mask 取消 pad 和句子a部分预测的影响
return (loss(predictions, labels) *
target_mask).sum() / target_mask.sum()
def forward(self,
input_tensor,
token_type_id,
position_enc=None,
labels=None,
device="cpu"):
# 传入输入，位置编码，token type id ，还有句子a 和句子b的长度，注意都是传入一个batch数据
# 传入的几个值，在seq2seq 的batch iter 函数里面都可以返回
input_shape = input_tensor.shape
batch_size = input_shape[0]
seq_len = input_shape[1]
# 构建特殊的mask
ones = torch.ones((1, 1, seq_len, seq_len),
dtype=torch.float32,
device=device)
a_mask = ones.tril() # 下三角矩阵
s_ex12 = token_type_id.unsqueeze(1).unsqueeze(2).float()
s_ex13 = token_type_id.unsqueeze(1).unsqueeze(3).float()
a_mask = (1.0 - s_ex12) * (1.0 - s_ex13) + s_ex13 * a_mask
enc_layers, _ = self.bert(input_tensor,
position_ids=position_enc,
token_type_ids=token_type_id,
attention_mask=a_mask,
output_all_encoded_layers=True)
squence_out = enc_layers[-1] # 取出来最后一层输出
predictions = self.decoder(squence_out)
if labels is not None:
# 计算loss
# 需要构建特殊的输出mask 才能计算正确的loss
# 预测的值不用取最后sep符号的结果 因此是到-1
predictions = predictions[:, :-1].contiguous()
target_mask = token_type_id[:, 1:].contiguous()
loss = self.compute_loss(predictions, labels, target_mask)
return predictions, loss
else:
return predictions
def generate(self,
text,
out_max_length=80,
beam_size=1,
device="cpu",
is_poem=False):
# 对 一个 句子生成相应的结果
# 通过输出最大长度得到输入的最大长度，这里问题不大，如果超过最大长度会进行截断
self.out_max_length = out_max_length
input_max_length = max_length - out_max_length
# print(text)
token_ids, token_type_ids = self.tokenizer.encode(
text, max_length=input_max_length)
token_ids = torch.tensor(token_ids, device=device).view(1, -1)
token_type_ids = torch.tensor(token_type_ids,
device=device).view(1, -1)
if is_poem: # 古诗的beam-search稍有不同
out_puts_ids = self.poem_beam_search(token_ids,
token_type_ids,
self.word2ix,
beam_size=beam_size,
device=device)
else:
out_puts_ids = self.beam_search(token_ids,
token_type_ids,
self.word2ix,
beam_size=beam_size,
device=device)
# 解码 得到相应输出
return self.tokenizer.decode(out_puts_ids)
def poem_beam_search(self,
token_ids,
token_type_ids,
word2ix,
beam_size=1,
device="cpu"):
"""
专门针对写诗的beam-search
"""
ix2word = {v: k for k, v in word2ix.items()}
sep_id = word2ix["[SEP]"]
douhao_id = word2ix["，"] # 逗号
juhao_id = word2ix["。"] # 句号
# 用来保存输出序列
output_ids = [[]]
word_list = {} # 保证不重复生成
last_chars = []
yayun_save = -1
# 用来保存累计得分
output_scores = torch.zeros(token_ids.shape[0], device=device)
flag = 0 # 判断第一次遇到逗号
for step in range(self.out_max_length):
scores = self.forward(token_ids, token_type_ids, device=device)
if step == 0:
# 重复beam-size次 输入ids
token_ids = token_ids.view(1, -1).repeat(beam_size, 1)
token_type_ids = token_type_ids.view(1,
-1).repeat(beam_size, 1)
# 计算log 分值 (beam_size, vocab_size)
logit_score = torch.log_softmax(scores, dim=-1)[:, -1]
logit_score = output_scores.view(-1, 1) + logit_score # 累计得分
# 取topk的时候我们是展平了然后再去调用topk函数
# 展平
logit_score = logit_score.view(-1)
hype_score, hype_pos = torch.topk(logit_score, beam_size)
indice1 = hype_pos / scores.shape[-1] # 行索引
indice2 = hype_pos % scores.shape[-1] # 列索引
# 下面需要更新一下输出了
new_hype_scores = []
new_hype_ids = []
next_chars = [] # 用来保存新预测出来的一个字符，继续接到输入序列后面，再去预测新字符
index = 0
for i_1, i_2, score in zip(indice1, indice2, hype_score):
i_1 = i_1.item()
i_2 = i_2.item()
score = score.item()
if i_2 != douhao_id and i_2 != juhao_id:
if i_2 not in word_list.keys():
word_list[i_2] = 1
else:
# 加惩罚
word_list[i_2] += 1
score -= 1 * word_list[i_2]
hype_score[index] -= 1 * word_list[i_2]
if flag == 0 and i_2 == douhao_id and len(last_chars) > index:
flag += 1
word = ix2word[last_chars[index]] # 找到上一个字符 记住其押韵情况
for i, each_yayun in enumerate(yayun_list):
if word in each_yayun:
yayun_save = i
break
if i_2 == juhao_id and len(last_chars) > index:
word = ix2word[last_chars[index]]
# 找押韵 给奖励
if word in yayun_list[yayun_save]:
score += 5
hype_score[index] += 5
else:
score -= 2
hype_score[index] -= 2
hype_id = output_ids[i_1] + [i_2] # 保存所有输出的序列，而不仅仅是新预测的单个字符
if i_2 == sep_id:
# 说明解码到最后了
if score == torch.max(hype_score).item():
return hype_id[:-1]
else:
# 完成一个解码了，但这个解码得分并不是最高，因此的话需要跳过这个序列
beam_size -= 1
else:
new_hype_ids.append(hype_id)
new_hype_scores.append(score)
next_chars.append(i_2) # 收集一下，需要连接到当前的输入序列之后
index += 1
output_ids = new_hype_ids
last_chars = next_chars.copy() # 记录一下上一个字符
output_scores = torch.tensor(new_hype_scores,
dtype=torch.float32,
device=device)
# 现在需要重新构造输入数据了，用上一次输入连接上这次新输出的字符，再输入bert中预测新字符
# 截取，因为要过滤掉已经完成预测的序列
token_ids = token_ids[:len(output_ids)].contiguous()
token_type_ids = token_type_ids[:len(output_ids)].contiguous()
next_chars = torch.tensor(next_chars,
dtype=torch.long,
device=device).view(-1, 1)
next_token_type_ids = torch.ones_like(next_chars, device=device)
# 连接
token_ids = torch.cat((token_ids, next_chars), dim=1)
token_type_ids = torch.cat((token_type_ids, next_token_type_ids),
dim=1)
if beam_size < 1:
break
# 如果达到最大长度的话 直接把得分最高的输出序列返回把
return output_ids[output_scores.argmax().item()]
def beam_search(self,
token_ids,
token_type_ids,
word2ix,
beam_size=1,
device="cpu"):
"""
beam-search操作
"""
sep_id = word2ix["[SEP]"]
# 用来保存输出序列
output_ids = [[]]
# 用来保存累计得分
output_scores = torch.zeros(token_ids.shape[0], device=device)
for step in range(self.out_max_length):
scores = self.forward(token_ids, token_type_ids, device=device)
if step == 0:
# 重复beam-size次 输入ids
token_ids = token_ids.view(1, -1).repeat(beam_size, 1)
token_type_ids = token_type_ids.view(1,
-1).repeat(beam_size, 1)
# 计算log 分值 (beam_size, vocab_size)
logit_score = torch.log_softmax(scores, dim=-1)[:, -1]
logit_score = output_scores.view(-1, 1) + logit_score # 累计得分
# 取topk的时候我们是展平了然后再去调用topk函数
# 展平
logit_score = logit_score.view(-1)
hype_score, hype_pos = torch.topk(logit_score, beam_size)
indice1 = hype_pos / scores.shape[-1] # 行索引
indice2 = hype_pos % scores.shape[-1] # 列索引
# 下面需要更新一下输出了
new_hype_scores = []
new_hype_ids = []
# 为啥有这个[],就是因为要过滤掉结束的序列。
next_chars = [] # 用来保存新预测出来的一个字符，继续接到输入序列后面，再去预测新字符
for i_1, i_2, score in zip(indice1, indice2, hype_score):
i_1 = i_1.item()
i_2 = i_2.item()
score = score.item()
hype_id = output_ids[i_1] + [i_2] # 保存所有输出的序列，而不仅仅是新预测的单个字符
if i_2 == sep_id:
# 说明解码到最后了
if score == torch.max(hype_score).item():
# 说明找到得分最大的那个序列了 直接返回即可
return hype_id[:-1]
else:
# 完成一个解码了，但这个解码得分并不是最高，因此的话需要跳过这个序列
beam_size -= 1
else:
new_hype_ids.append(hype_id)
new_hype_scores.append(score)
next_chars.append(i_2) # 收集一下，需要连接到当前的输入序列之后
output_ids = new_hype_ids
output_scores = torch.tensor(new_hype_scores,
dtype=torch.float32,
device=device)
# 现在需要重新构造输入数据了，用上一次输入连接上这次新输出的字符，再输入bert中预测新字符
# 截取，因为要过滤掉已经完成预测的序列
token_ids = token_ids[:len(output_ids)].contiguous()
token_type_ids = token_type_ids[:len(output_ids)].contiguous()
next_chars = torch.tensor(next_chars,
dtype=torch.long,
device=device).view(-1, 1)
next_token_type_ids = torch.ones_like(next_chars, device=device)
# 连接
token_ids = torch.cat((token_ids, next_chars), dim=1)
token_type_ids = torch.cat((token_type_ids, next_token_type_ids),
dim=1)
if beam_size < 1:
break
# 如果达到最大长度的话 直接把得分最高的输出序列返回把
return output_ids[output_scores.argmax().item()]