Helsinki-NLP OPUS-MT-train issues

Helsinki-NLP / OPUS-MT-train

Training open neural machine translation models

MIT License

318 stars 40 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

how to control maximum length

#106 jemesome opened 1 month ago
1
Wrong repetitive phrases for chinese to english translation

#105 thisisandreeeee opened 1 month ago
1
question about SPM and BPE

#104 jemesome closed 1 month ago
1
Training dataset statistics or technical reports

#103 phphuc612 closed 3 weeks ago
1
zh-en model specs

#102 franckbrl opened 3 months ago
0
Decode target & source smp

#101 nichellehouston opened 3 months ago
0
No Latin models?

#100 d3287t328 opened 4 months ago
0
New model

#99 Rom4rio74 opened 5 months ago
0
What's the prefix for "Helsinki-NLP/opus-mt-zh-en" when using "AutoModelForSeq2SeqLM"?

#98 blackli7 opened 6 months ago
0
I want a ja-ko model.

#97 sawook-min opened 7 months ago
0
opus-mt-tc-big-tr-en doesn't work properly.

#96 bleriot14 opened 9 months ago
0
Fine-tuning models using HuggingFace libraries

#95 blademoon opened 10 months ago
0
Fine-tuning with guided alignment

#94 Kai-Piontek opened 1 year ago
0
preprocess.sh [: ==: unary operator expected

#93 thammegowda closed 1 year ago
2
Lack of transparency on used training data. - Does finetuning make sense?

#92 Thybo-D opened 1 year ago
1
Data for Brazilian Portuguese

#91 reletreby opened 1 year ago
2
Unbelievably High BLEU scores from finetuning...

#90 hdeval1 closed 1 year ago
3
how to train our dataset

#89 bondeakshay opened 1 year ago
3
Hyperparameters used for pretrained models?

#88 davisyoshida opened 1 year ago
1
Unable to find current origin/master revision in submodule path

#87 hthanhbmt opened 1 year ago
2
Reproduced crash on Opus-mt-en-de model using string "J" and "J-10"

#86 Qubitium closed 1 year ago
1
different sizes of dictionaries in different models

#85 bariluz93 opened 1 year ago
1
update Dockerfile.gpu--fixed

#84 gotomypc opened 1 year ago
0
Using OPUS-MT with DeepSpeed

#83 rlenain opened 1 year ago
0
How to translate from english to Japan?

#82 izzahzawave closed 1 year ago
0
Wrong tokenizer/vocab for the 'Helsinki-NLP/opus-mt-tc-big-en-ko' model

#81 regpath opened 1 year ago
0
What could cause widely varying inference time when using pre-trained opus-mt-en-fr model with python transformers library?

#80 shandou opened 1 year ago
2
Multilingual Tuned Model Translating everything to "sssssssss"

#79 hdeval1 opened 1 year ago
2
Korean Finetuning

#78 hdeval1 opened 2 years ago
0
Preprocessing Script Question

#77 hdeval1 opened 2 years ago
0
What is tatoeba-langtune?

#76 hdeval1 closed 2 years ago
2
Language Code Difference

#75 hdeval1 closed 2 years ago
1
What's the dataset used for training opus-mt-en-de

#74 JiayiFeng opened 2 years ago
1
Bridge Language

#73 seekingdeep closed 2 years ago
0
Source.spm & Target.spm Files

#72 hdeval1 closed 2 years ago
2
Preparing fine-tune data for Marian

#71 JoeTseHot opened 2 years ago
2
Model not available on huggingface model page, how do I use it with huggingface.

#70 mt-empty opened 2 years ago
4
HuggingFace conversion script doesn't work

#69 mcPear opened 2 years ago
1
Preprocessing of training data

#68 alexvs-sysoev opened 2 years ago
1
Problem Fine-tuning Models using TMX files

#67 hdeval1 closed 2 years ago
5
model Helsinki-NLP/opus-mt-en-uk translates some sentences into Russian instead of Ukrainian

#66 shyrma opened 2 years ago
1
how to fine tune the model published on huggingface | opus-mt-en-zh

#65 QzzIsCoding opened 2 years ago
1
Patch 1

#64 rrrepsac closed 1 year ago
0
new Huggingface model Helsinki-NLP/opus-mt-fr-it

#63 JoelNiklaus opened 2 years ago
5
broken translation, many repetitions & duplicates

#62 kindziora opened 2 years ago
1
Request for model Helsinki-NLP/opus-mt-en-ko

#61 linlinsam opened 2 years ago
2
Offline translation?

#60 zubairahmed-ai opened 3 years ago
1
Question: Helsinki-NLP/opus-mt-fi-fi ??

#59 remotejob opened 3 years ago
1
Dataset used for training the models

#58 maroxtn closed 3 years ago
2
Marian compilation: max 8 jobs to avoid memory error

#57 raphaelmerx closed 3 years ago
0