EasyNMT
EasyNMT copied to clipboard
UKPLab
Easy to use, state-of-the-art Neural Machine Translation for 100+ languages
I really like the library, great work! Is there a way to manually specify a specific OPUS model? For example EasyNMT with OPUS currently does not support English as source...
Thanks a lot for creating this great package! Question: will every translation with equivalent input always lead to the exact same output, or is there some randomness involved (e.g. through...
Hi, Thank you for this useful library. I tried to install in my machine and it gave me this error. Any help please? print(model.translate('This is a sentence we want to...
easynmt docker install is working fine through http requests. Now i'd like to run some benchmark. How do you run /examples/test_translation_speed.py ?
Hi! I was wondering if there is a workflow for large datasets available. I am trying to translate a big amount of tweets using Pandas and Python. Best, Daniel
Exception when trying to download http://easynmt.net/models/v2/opus-mt/easynmt.json. Response 403 im so sad about this
Bug fix - max_length parameter was not being propagated to AutoModel tokenizers. This was causing OOM for my system on large document inputs.
  OSError: Unable to load weights from pytorch checkpoint file for 'facebook/m2m100_1.2B' at '/cache/transformers/68002fb1a7773d8d8373f1a230588141964ef9f249db6987681f295dbe85356c.ee70663869b89be4f68eed03a21d5c3400b223cb544883f411e469aaea0a25f9'If you tried to load a PyTorch model from a TF 2.0 checkpoint, please set from_tf=True.
With the [switch away from Fairseq to huggingface transformers](https://github.com/UKPLab/EasyNMT/commit/9b69703af901a7aeb83b5dda0830123a91c307ab) in EasyNMT 2.0, there seem to have been substantial changes, e.g. the size of the `m2m_100_1.2B` model has [increased from 2.3...
