Issues
- 2
- 3
Further standardize the layout and packaging
#271 opened by mjpost - 4
no published sdist for 2.4.2
#269 opened by dhellmann - 2
- 1
BLEU and chrF2 on Chinese
#240 opened by zhhl9101 - 2
Add BLEU max_ngram_order to signature
#251 opened by BramVanroy - 0
Memory leak in spm tokenizers
#264 opened by nkrasner - 1
Difference between cli and corpus_score
#259 opened by VarunGumma - 1
when use package 'evaluate‘ with 'sacrebleu' calculate metric happend error
#263 opened by TristanShao - 4
No references found for test set wmt23/*
#261 opened by kellymarchisio - 0
__init__.py divergence in 2.4.0
#255 opened by dustalov - 2
Compute BLEU from Python mishandles references
#256 opened by OrianeN - 4
Accelerating sentence scores
#254 opened by AmitMY - 4
Problems about the tokenizers
#192 opened by DoctorDream - 1
Add WMT 23 test sets
#245 opened by mjpost - 1
[Feature Request] HTER Implementation
#248 opened by shivanraptor - 3
Taking a long time to download the test set
#242 opened by Phuoc-Hoan-Le - 2
Schedule a new release
#241 opened by daskol - 1
What's the difference between setting "--tokenize" to "flores101" and setting it to "flores200"?
#243 opened by Phuoc-Hoan-Le - 1
Working on tokenized pairs?
#244 opened by MostHumble - 0
- 0
- 1
TER asian support
#229 opened by esalesky - 2
Discrepancy in docstrings in TER `normalized`
#230 opened by BramVanroy - 3
TER between two empty strings is 100
#228 opened by BramVanroy - 2
GitHub workflows CI tests fail on Python 3.6
#233 opened by martinpopel - 0
- 0
Remove the flores101 related tokenizer logging message when the selected tokenizer is flores200
#219 opened by hadyelsahar - 1
Unicode normalization?
#224 opened by davidweichiang - 0
- 4
Default params for calculating bleu
#195 opened by base-y - 2
Calculate sentence-level and corpus-level BLEU with tokenizer flores101(or flores200) on GPU
#227 opened by ElizabethUniverse - 1
Empty lines in WMT21/dev Icelandic-English
#225 opened by ZJaume - 2
module 'sacrebleu' has no attribute 'corpus_bleu
#222 opened by gongel - 12
TER above 100?
#208 opened by JoyeBright - 1
lru_cache max size is not set in tokenizer_spm.py
#217 opened by chikiulo - 0
Add WMT22 data
#215 opened by mjpost - 4
Switch to poetry for build/install
#202 opened by mjpost - 8
Installing latest version via PIP results in an error
#209 opened by epwalsh - 0
Support for printing significance score as JSON
#207 opened by me-manikanta - 2
List test sets available for a given language pair
#210 opened by ZJaume - 0
Support test sets from sign language task?
#212 opened by bricksdont - 1
Incorrect sample size?
#206 opened by kocmitom - 2
How to calculate the chrf2 score?
#193 opened by RamoramaInteractive - 3
[Feature Request] Word Error Rate
#199 opened by jonathanmutal - 1
tarball is not defined
#201 opened by VanyaBK - 3
Why give 0 bleu score when evaluating De ?
#198 opened by Hannibal046 - 2
more accurate exception types on argument checks
#189 opened by abcdenis - 0
- 2
Is ‘extract_ngrams’ removed in the latest version?
#188 opened by rangehow