issues
search
yixinL7
/
BRIO
ACL 2022: BRIO: Bringing Order to Abstractive Summarization
327
stars
43
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
> Hi,I'd like to ask how long it takes you to train an epoch.Thanks
#38
hoboyu11
closed
7 months ago
0
If can process long Chinese data
#37
991126
opened
1 year ago
1
Conflicting between mle and ranking loss during training
#36
thangld201
opened
1 year ago
1
Unable to generate summary when initializing model with PyTorch
#35
yungsinatra0
opened
1 year ago
1
I was wondering what was max_len, total_len was calculated is it by words or by each characters?
#34
mrasyadc
opened
1 year ago
0
environment.yml needed
#33
mrasyadc
closed
1 year ago
0
Question about training loss settings
#32
Runingtime
closed
1 year ago
1
Can't load tokenizer using from_pretrained, please update its configuration: No such file or directory (os error 2)
#31
Alsace08
opened
1 year ago
0
Converting BRIO saved checkpoints to huggingface format
#30
tanay2001
closed
1 year ago
1
About data loading
#29
ZhaoQingAn
opened
1 year ago
0
How to calculate BERTScore in the paper
#28
cq-cdy
opened
1 year ago
0
1. I want to know whether "Preprocessed Data" is generated through "Generate Candidate Summaries". 2. If "Own Data" is processed through preprocess.py, then the test.source, test.source.tokenized and other files in src_dir Where did it come from, was it created manually?
#27
cq-cdy
opened
1 year ago
1
Running too slow when using gen_candidate.py to generate candidates
#26
aidejieceng
opened
1 year ago
1
About cased cnn daily mail dataset
#25
ruili33
opened
1 year ago
0
Using BRIO for text summarization on another language
#24
thaokimctu
closed
1 year ago
3
data preprocessing
#23
xjw-star
opened
1 year ago
3
Dataset issue
#22
sarahwie
opened
1 year ago
1
Model checkpoints for nyt
#21
tagoyal
opened
1 year ago
2
Create raw_data structure in example folder from input
#20
tungphamMTA
opened
1 year ago
1
Params gold_weight is always 0?
#19
lxx909546478
closed
1 year ago
2
There is a bug in the Preprocessed cnndm dataset
#18
aidejieceng
closed
1 year ago
4
About the scoring mode
#17
ZhouxiangFang
opened
2 years ago
1
Ranking Loss Question
#16
griff4692
closed
2 years ago
4
Is it possible to train Huggingface model using BRIO class?
#15
dleunji
opened
2 years ago
1
Specified model filename pattern was: #ID#.ref
#14
caicai83121
opened
2 years ago
1
About training time
#13
EnghishYang
opened
2 years ago
2
Update README.md
#12
Buckeyes2019
opened
2 years ago
0
RuntimeError: CUDA out of memory when running training command
#11
tiennvcs
closed
2 years ago
1
A small trick for memory efficiency
#10
Hannibal046
opened
2 years ago
1
Apply BRIO to other generation tasks
#9
HillZhang1999
opened
2 years ago
8
Hard coded max_length, it should be `args.total_len`.
#8
CheaSim
closed
2 years ago
1
GPU usage increasing as training progresses
#7
thechargedneutron
opened
2 years ago
7
MIT lincense?
#6
seungeunrho
closed
2 years ago
2
questions about ROUGE score
#5
jeewoo1025
closed
2 years ago
2
Modified gen_candidate.py
#4
jeewoo1025
closed
2 years ago
2
Why do lower case for BARTCNN?
#3
Ricardokevins
closed
2 years ago
2
Why remove embedding for <mask> in BART?
#2
FeiWang96
closed
2 years ago
0
HuggingFace Tokenizer Loading
#1
griff4692
closed
2 years ago
1