v1.0.1

minor fix

v1

• fine-tuning results

The fine-tuning results when title_idx is 0 - 10.

{'accuracy': 0.7660256410256411, 'nmi': 0.8386512061608621}
{'accuracy': 0.7093023255813954, 'nmi': 0.6706178595371249}
{'accuracy': 0.8909395973154363, 'nmi': 0.8648364310117125}
{'accuracy': 0.5889328063241107, 'nmi': 0.7376563301923733}
{'accuracy': 0.5787401574803149, 'nmi': 0.6425792338794517}
{'accuracy': 0.8435619735258725, 'nmi': 0.8173820198288065}
{'accuracy': 0.6141367323290846, 'nmi': 0.6476360662568814}
{'accuracy': 0.6615661566156615, 'nmi': 0.6160746018562953}
{'accuracy': 0.3350202429149798, 'nmi': 0.5890785182885055}
{'accuracy': 0.6675257731958762, 'nmi': 0.7386078079778033}
{'accuracy': 0.6675257731958762, 'nmi': 0.7386078079778033}

pre

• pre-training code
• pre-trained model
• fine-tuning code

We reported the average score over five trials of clustering in our README.
We list one of the results below.

                     accuracy       nmi
ARMS                 0.621795  0.722185
AisazuNihaIrarenai   0.446512  0.548274
AkkeraKanjinchou     0.605705  0.705397
Akuhamu              0.363636  0.614287
AosugiruHaru         0.407480  0.526837
AppareKappore        0.575211  0.698233
Arisa                0.475087  0.618468
BEMADER_P            0.512151  0.501424
BakuretsuKungFuGirl  0.274291  0.527831
Belmondo             0.526252  0.671235
LoveHina_vol14       0.438144  0.633914
accuracy    0.476933
nmi         0.615280