bert_score

bert_score copied to clipboard

I am trying to compare each of a large number of candidate sentences (N = 28,830) to a list of the same reference sentences (N = 382) and get the...

tierney6

Calculate embeddings once

1

Hello! I did not find Discussions for the repository, so I will question here. I am wanted a high level interface, but a bit lower: for calculating embeddings and getting...

ivanstepanovftw

Hi there. Is there any way to run bert_score on tpu? If yes, please explain or send a straightforward link. I'd like to use the score function like below: score(input_txt,...

seyyedjavadrazavi

Dear sir/madam, Could you please clarify/guide me how to incorporate covid-twitter-bert model ( https://huggingface.co/digitalepidemiologylab/covid-twitter-bert-v2 ) with bert_score. I did not find this model in the suggested googlesheet of the supported...

volhakatebi

using torch==1.5.1 using transformers==3.0.1 using bert-score==0.3.4 P, R, F1 = score(cands, refs, lang='hi', model_type='bert-base-multilingual-cased') Above line of code is giving the below error. I have tried transformer version 4.5.1 and...

agrawalanshul053

Could you please add some models for Ukrainian language?

4

Currently I know a few: https://huggingface.co/youscan/ukr-roberta-base https://huggingface.co/dbmdz/electra-base-ukrainian-cased-discriminator https://huggingface.co/lang-uk/electra-base-ukrainian-cased-discriminator https://huggingface.co/lang-uk/electra-base-ukrainian-dbmdz-vocab-cased-discriminator

dchaplinsky

will bert-score cli command automatically use gpu? how?

1

Dlinuigh

get_wmt18_seg_result question

3

您好，我在运行get_wmt18_seg_result.py文件时，得到的如下结果，这和您在论文中报告的absolute pearson correlation 并不相同。 您可以解释下get_wmt18_seg_result.py文件得到的结果吗？ 我应该怎么样得到论文中报告的absolute pearson correlation结果呢？ model_type cs-en de-en et-en fi-en ru-en tr-en zh-en avg roberta-large P 0.3831702544031311 0.54582257007364 0.39514465541862803 0.2939672801635992 0.35140330642060746 0.29337243401759533 0.2487933567167311 0.35881055103056175 roberta-large R 0.40626223091976515...

Hou-jing

Mabey we cannot assert 'model_type' as a local path?

6

141forever

load local models for bert-score. just change `model_type` to your local model path, e.g. ```python self.bert_scorer = BERTScorer( model_type="/Users/wangyangfan/remote/models/FacebookAI/roberta-large", device='mps', rescale_with_baseline=True, lang='en' ) ``` 👏 Inspired by @shuaijiumei, thanks a...

WangYangfan