Datasets:

TingChen-ppmc
/

Shanghai_Dialect_Conversational_Speech_Corpus

Modalities:

Audio

Text

Formats:

Size:

Libraries:

Dataset card Viewer Files Files and versions Community

Dataset Viewer

Auto-converted to Parquet

Full Screen Viewer

Full Screen

Split (1)

train · 3.79k rows

gender stringclasses 2 values	speaker_id stringclasses 20 values	transcription stringlengths 1 62
male	G0003	北京爱数智慧语音采集
female	G0004	北京爱数智慧语音采集
male	G0003	阿拉两个拧来聊聊金融方面呃
male	G0003	金融方面嘛
male	G0003	搿呃，阿姨喃，应该讲，侬已经交关年数辣辣了解了
male	G0003	葛末，吾辣辣金融方面已经有的三四年了
male	G0003	最少辰光阿拉是做撒呃喃，有钞票就是到银行里保本保息
female	G0004	吾已经做了已经到八七年了
female	G0004	八七年呃，当时辰光辣里哴相做理财呃
female	G0004	是两级风险，三级风险，四级风险吾侪做呃，信托咯撒侪做呃
female	G0004	信托有两年头呃八点几
female	G0004	吾也做呃，尝试呃
male	G0003	但是呢，侬讲呃种信托啊
male	G0003	呃
male	G0003	过去做呃信托呢，是保本保息呃
male	G0003	国家辣辣银行里兜呃
male	G0003	现在呃信托啦，假呃啦
male	G0003	讲呃撒呃信托，撒呃保本保息，现在危险了，撒道理呢，信托亏比侬现在金融喃
male	G0003	侪拨国家拦特了，撒叫拦特了
male	G0003	本来动，比方侬是信托，本来是银行是保，担保呃
male	G0003	现在银行伐担保了
male	G0003	国，国家有个先，有个文件就讲
male	G0003	银行里如果破产超过五十万以上呃
male	G0003	国家伐付钞票呃
male	G0003	超过五十万以内呃，国家负担搿钞票，搿是正式文件谈下来呃
male	G0003	葛末，阿拉辣辣搿问题高头呢，阿拉辣搿问题高头呢，就是讲哪能来认识呢
female	G0004	现在对伐，老早金融行业呢，本身就是伐保本呃，也伐保本呃所谓五十万赔偿呃损失，是侬做是撒
female	G0004	叫撒是
female	G0004	呃，呃银行理财，银行呃产品
female	G0004	是，就是坏特了，伊拨侬五十万
female	G0004	呃叫撒赔拨侬呃，但是做金融呃行业，本身也就伐赔侬呃
female	G0004	现在呢就是有得一只，老早是伐赔喏，老早讲起来呢，坏呢，就是讲风险呢，但是呢
female	G0004	还可以呃，为撒道理，基本上没呢没撒坏呃
female	G0004	但是现在呢，是刚性兑付了，新呃文件出来是刚性兑付，所以讲万一搿只物事没做好呃闲话
male	G0003	还有，关于讲吾，讲到搿呃就是讲信托
male	G0003	银行呃理财
male	G0003	呃，吾觉着呢，侬讲了蛮有道理呃
male	G0003	银行理财产产品呢是辣辣五十万里相头
male	G0003	对伐，伊可以保保周梦伐
male	G0003	超过五十万伊伐保呃
male	G0003	但是，吾今朝再晓得，侬搿能噶讲呢吾理解，吾本来想呢
male	G0003	放侬银行里钞票
male	G0003	五十万以下伊拨侬保呃，五十万以上伊伐保呃，哦，原来是银行里呃理财产品
male	G0003	可以保呃
male	G0003	超过五十万，伊伐保呃，是伐是啊
female	G0004	王老师，侬侬搞错特了，银行里相五十万是等于存款
female	G0004	侬拨辣存款，就是讲一年头呃，或者是半年头，存款，万一搿只银行坏特了，伊是赔拨侬呃
female	G0004	但是银行呃所有理财产品侪伐保呃
male	G0003	哦，是搿能噶，哦[+]
male	G0003	哦，搿是，侬搿能样子跟吾讲了蛮清爽呃
male	G0003	乃末，还有一个呢，最近呢就是讲吾有两个朋友呢叫吾搿是，参加搿呃就是讲
male	G0003	呃，就是理财
male	G0003	搿呃公司啊去参加搿理财活动
male	G0003	有点理财公司呢
male	G0003	讲讲是阿拉侪是规范呃撒呃证照侪齐呃
male	G0003	到后头来，伊讲，阿拉搿搭要求投资拧要投一百万
male	G0003	投二百万，投五百万
male	G0003	后手吾问伊拉，侬到底侬是只撒公司
male	G0003	做撒物事
male	G0003	甚至谈到后头来侬投三万，投十万，廿万，三万也可以投呃
male	G0003	吾觉着伊搿只公司
male	G0003	可能怀疑性，是伐是正规呃侬讲讲呢，吾听听侬呃
female	G0004	王老师，吾是搿能噶想呃，因为是辣辣一两年辰光刚刚普及了金融理财理财搿只物事
female	G0004	阿拉呢去听听课，本身呢带辣一只撒心情呢，去看看叫，了解一下
female	G0004	结果了解了以后嘛，天天上课啊，讲特一歇通货膨胀搿点物事呢，听法听法呢，觉着倒还还可以嘛
female	G0004	阿拉应该呢，就是讲也参加一些，葛末吾认为呢，邓小平讲了一句闲话到蛮对呃
female	G0004	胆子呢要大一点步子呢要快一点
female	G0004	葛末，由于伊搿只想法呢，倒蛮符合吾要求呃，诶，吾觉着是应该，既然侬要想钱生钱
female	G0004	别过胆子要大一点，葛末吾就开始呢投进去了，但是因为当时辰光侪伐晓得是要坏特呃
female	G0004	葛末辣辣搿种情况下头，有的担保啊，有的营业执照啊，有的撒物事啊，阿拉侪比较相信呃，进去了
female	G0004	现在呢，坏呃蛮多呃特别辣辣一一五年呃上半年
female	G0004	一六年呃下，哦一五年呃
female	G0004	上，下半年，一六年呃上半年跑路呃老多呃
female	G0004	所以呢，现在跑出来搿种公司咯撒
female	G0004	吾觉着是老伐靠谱呃因为阿拉侪有的阴影了，侪坏特过了
female	G0004	所以呢，现在再去搿种呢
female	G0004	再去哪能噶就伐大会的轻易呃去相信人家了
female	G0004	王老师侬讲对伐
male	G0003	侬讲了蛮有道理呃
male	G0003	因为过去呃，去年伐前年
male	G0003	嗯辣社会高头讲P to P
female	G0004	埃能噶再出去投呢，就要警觉了，因为有的坏过了
female	G0004	人家讲，一次拨蛇咬，十年要怕惊蛇了，就讲现在呢要小心点了
female	G0004	老是侵犯人家讲好撒物事，有执照辣，有撒物事，因为伊伐讲好，那哪能会的进去呢
female	G0004	伊如果实事求是讲，阿拉里相已经亏损了，那来帮帮忙，撒拧还会的去帮忙
female	G0004	只因为伊侪讲了好，侬相信了再去投
female	G0004	吾觉着现在出来呃公司呢，侪老伐靠谱呃，为撒道理，现在伐是，像老早搿能噶一万二万或者五万门楷
female	G0004	十万就可以进去了，现在狮子大开口，要一百万，六百万
female	G0004	人家就是有的一百万，六百万，人家也伐会的进去呃，但是现在交关拧已经受过伤了
female	G0004	侪坏特过了，侬讲像现在提出搿种一百万基金咯撒，人家哪能会的进去呢
female	G0004	老早吾记得辣辣一四年辰光，吾做一只基金，伊也要一百万，但是呢，伊要交关拧凑起来
female	G0004	大家侬凑十万，吾凑十万，并呃一百万，有公司高那去做基金呃，但是现在是要一个人一百万
female	G0004	吾心里相一只辣想[+]
female	G0004	搿一百万，就是现在老百姓已经坏特交关了，也伐可能去拿出一百万来，必定有几个拧
female	G0004	有几家人家有的一百万，又伐是做生意拧
female	G0004	如果现在老板做生意呃吾相信也伐会的投一百万
female	G0004	伊是撒，三加五，所以吾，所谓三加五就是三年加两年，相当于要五年
female	G0004	搿五年钞票扔进去等于是基本哴是伐好动呃
female	G0004	葛末侬作为一个老板，肯定要资金要周转呃
female	G0004	伊有的一百万，身边最起码要有的四百万伐

End of preview. Expand in Dataset Viewer.

Corpus

This dataset is built from Magicdata ASR-CZDIACSC: A CHINESE SHANGHAI DIALECT CONVERSATIONAL SPEECH CORPUS

This corpus is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. Please refer to the license for further information.

Modifications: The audio is split in sentences based on the time span on the transcription file. Sentences that span less than 1 second is discarded. Topics of conversation is removed.

Usage

To load this dataset, use

from datasets import load_dataset
dialect_corpus = load_dataset("TingChen-ppmc/Shanghai_Dialect_Conversational_Speech_Corpus")

This dataset only has train split. To split out a test split, use

from datasets import load_dataset
train_split = load_dataset("TingChen-ppmc/Shanghai_Dialect_Conversational_Speech_Corpus", split="train")
# where test_size=0.5 denotes 0.5 of the dataset will be split to test split
corpus = train_split.train_test_split(test_size=0.5)

A sample data would be

# note this data is from the Nanchang Dialect corpus, the data format is shared
{'audio': 
    {'path': 'A0001_S001_0_G0001_0.WAV',
       'array': array([-0.00030518, -0.00039673, 
                     -0.00036621, ..., 	-0.00064087,
                      -0.00015259, -0.00042725]),
       'sampling_rate': 16000},
 'gender': '女',
 'speaker_id': 'G0001',
 'transcription': '北京爱数智慧语音采集'
}

More Information needed

Downloads last month: 36

Edit dataset card

Size of downloaded dataset files:

437 MB

Size of the auto-converted Parquet files:

437 MB

Number of rows:

3,792

Models trained or fine-tuned on TingChen-ppmc/Shanghai_Dialect_Conversational_Speech_Corpus

TingChen-ppmc/whisper-small-Shanghai

Automatic Speech Recognition • Updated Dec 20, 2023 • 9