bert_score
bert_score copied to clipboard
BERT score for text generation
```python from bert_score import score with open("/data/translator/parallel/kaggle/open_ru.txt", "r") as fo: txts = [t.strip() for t in fo.readlines()] txts = list(set(txts)) size = 1_000_000 ``` ```python P, R, F1 = score(...
I just add some codes for using BERTScore on Apple Silicon and it works fine. I think MPS support should be add to the code base for Mac users' convenience....
If I use distilbert-base-uncased model trans_version 4.40 It will have max_length 1000000000000000019884624838656 in the utils.py line 216 DistilBertTokenizer(name_or_path='distilbert-base-uncased', vocab_size=30522, model_max_length=1000000000000000019884624838656, is_fast=False, padding_side='right', truncation_side='right', special_tokens={'unk_token': '[UNK]', 'sep_token': '[SEP]', 'pad_token': '[PAD]', 'cls_token':...
Greetings, I am using the BertScore for multilingual translation scoring. I have a source in English, a vendor provided translation, and a Spanish translation produced by Azure's AI Services. When...
Hi, I am going to use the clinical models like "**BSC-LT/roberta-base-biomedical-clinical-es**", "**Clinical BioBER**T", or even one of the clinical models in this link ["https://huggingface.co/models?search=clinical"](https://huggingface.co/models?search=clinical) as a model_type in bert score...
https://github.com/huggingface/transformers/issues/8739#issuecomment-732412022 Can you update the name of arguments according to newer transformers library? Thank you for reading this issue :)
maybe it's worth to load bert considering argument language. not just load bert cased
There is a bug in the `score()` method's code: A `batch_size` attribute can be specified for the `BERTScorer` object, but this setting is ignored by the `score()` method currently. In...
OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like roberta-large is not the path to a directory containing...