I want to write the sentences written in the tsv file with spaces and store the words in each sentence as nested structural lists.

 from gensim.models.doc2vec import Doc2Vec
from gensim.models.doc2vec import TaggedDocument

import MeCab
import csv

mt = MeCab.Tagger()

reports=[ ]# List of TaggedDocuments (remain original question)
wordDic={}#List of report_id for each line and only words per sentence

# Word of mouth ID and word of mouth separated by tab for each line
tsv=["1\tStart this LED with usb", "When did 2\tnasa and jaxa make ohashi", "3\t That ojigi's degrees are too big"]
reader=csv.reader(tsv, delimiter="\t")
for report_id, report in reader:
    words = [ ]
    node=mt.parseToNode(report)
    while node:
        if node.feature.split(",")[0]==u "noun":
            words.append(node.surface)
        elif node.feature.split(",")[0]==u "adjective":
            words.append(node.feature.split(",")[6])
        whale lice. featurename Sensinode Second Life. the splitmethod (" , " ) [ 0 ]   =   " verb " :
            words.append(node.feature.split(",")[6])
        node = node.next

    wordDic [report_id] = words

stopword = ['れる','の','られる','せる','させる','LED','pantone','20','30','β','3','？','9','（','\u3000','＆','A','W','％','×','"','６','15','1994','３',"'",'GW','25','OO','s','1','/','500','%','.','0','2020','V','13','5','180','-', '%)','(','+','４','8','10','100','14','ABS','4','K','"','７','50','．','33','５','上','器','丿','・','０','40','7', '?','）','60','＋','２','／','、','。','「','」','2','1000','1010123','1093','12','120','16','17','1930','1972','200','2002','21','24','260','27','34','35','36','360','37','40000','4136','43','480','70','75','agata','back','ball','basic','cd','chadwick','chair','co','crt','daa','degrees','dm','dvd','eco','epd','eva','fsc','gulfman','homearchi','hybrid','ic','ipad','ipod','lan','leavbes','led','may','md','mdf','mf','mm','mmx','ms','nanoe','nasa','nc','no','oa','off','ohashi','ojigi','on','pantone','pc','pet','pp','prism','pro','rom','sd','shock','soho','td','the','to','tpo','trek','usb','ventura','way','wear','ａｖ','ｄｉｙ','led','ｌｅｄ','ｐｃ','１つ','２つ','そう','どこ','ここ','これ','この','いつ','こんな','いくつ','これら','どの','その','それぞれ','そのもの','こうした','それ',' sore nari ', ' kochira ', ' one ', ' kou iu ', ' sou iu ', ' In these ta ', ' dou ', ' dochira ', ' dore ', ' how ', ' No matter how ', ' mono ', ' koto ', ' tokoro ', ' you ', ' tame ', ' yellowtail ', ' often ', ' bulk ', ' tightly-closed ', ' tou ', ' degree ', ' toki ', ' koto ',       ,       ,       ,       ,       ,       ,       ,       ,       ,       ,       ,       ,       ,える','かんがえる','わかる','分かる','分る','見える','みえる','知る','しる','知れる','しれる','言える','いえる','示す','しめす','述べる','のべる','書く','かく','よる','だす','出す','入る','はいる','いれる','入れる','使う','つかう','用いる','もちいる','持つ','もつ','もてる','持てる','作る','つくる','なす','起こる','おこる','つく','つける','付く','付ける','聞く','よぶ','きく','呼ぶ','ない','高い','低い','多い','少ない','強い','大きい','小さい','長い','ながい','よい','良い','悪い','いい','やすい','にくい','うWell bad for ' , ' design ' , ' ' , ' kinou ' , ' goods ' , ' products '] like ' , '
for report_id in wordDic:
    print(wordDic[report_id])
    words2 = [ token for token in wordDic [ report_id ] if token not in stopword ]
    reports.append(TaggedDocument(words=words2, tags=[report_id]))

print(reports)
# a code that lists only words for each sentence
print(list(wordDic.values()))
# a code that lists only words 2 for each sentence by sentence
print([report.words for report in reports])

['LED', 'usb', 'Start', 'Do', 'Let']
['nasa', 'jaxa', 'is', 'ohashi', 'make', 'of']
["ojigi", "degrees", "big", "too much"]
[TaggedDocument (words=['start'], tags=['1'], TaggedDocument (words=['jaxa'], tags=['2'], TaggedDocument (words=[], tags=['3']]]]
[[['LED', 'usb', 'start', 'do', 'let'], ['nasa', 'jaxa', 'is', 'ohashi', 'make', 'of', 'ojigi', 'degrees', 'big', 'too big']]]]
[[['Start', 'jaxa', []]