You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@opennlp.apache.org by Martin Wiesner <ma...@apache.org> on 2024/11/17 09:48:18 UTC

[DISCUSS] Release pre-trained OpenNLP Models (1.2) ?

Hi,

Any objections against (or support for) doing a release (1.2) for the pre-trained models? Details below:

I’ve added 9 new languages (via UD treebanks) to our model list:

- "Armenian|hy|BSUT"
- "Basque|eu|BDT"
- "Catalan|ca|AnCora"
- "Georgian|ka|GLC"
- "Greek|el|GDT"
- "Kazakh|kk|KTB"
- "Korean|ko|Kaist"
- "Icelandic|is|IcePaHC" 
- "Turkish|tr|BOUN"

In total, this results in 32 supported languages, each with sentence detection, tokenization and POS tagging.

Moreover, the updated ud-train script now produces ME models for the Lemmatizer component, 32 of them.

The training is conducted with OpenNLP 2.5.0 and with the treebanks from the latest UD release, dating Nov 15, 2024.

I can act as RM.

Best
Martin

Re: [DISCUSS] Release pre-trained OpenNLP Models (1.2) ?

Posted by Bruno Kinoshita <br...@gmail.com>.
+1! I can more or less test the Catalan, but at work there are two
colleagues from the Basque Country and I think I could check with them.

Thanks a lot!

On Sun, 17 Nov 2024 at 11:01, Richard Zowalla <ri...@zowalla.com> wrote:

> +1 (do it)
>
> Am 17. November 2024 10:48:18 MEZ schrieb Martin Wiesner <
> mawiesne@apache.org>:
> >Hi,
> >
> >Any objections against (or support for) doing a release (1.2) for the
> pre-trained models? Details below:
> >
> >I’ve added 9 new languages (via UD treebanks) to our model list:
> >
> >- "Armenian|hy|BSUT"
> >- "Basque|eu|BDT"
> >- "Catalan|ca|AnCora"
> >- "Georgian|ka|GLC"
> >- "Greek|el|GDT"
> >- "Kazakh|kk|KTB"
> >- "Korean|ko|Kaist"
> >- "Icelandic|is|IcePaHC"
> >- "Turkish|tr|BOUN"
> >
> >In total, this results in 32 supported languages, each with sentence
> detection, tokenization and POS tagging.
> >
> >Moreover, the updated ud-train script now produces ME models for the
> Lemmatizer component, 32 of them.
> >
> >The training is conducted with OpenNLP 2.5.0 and with the treebanks from
> the latest UD release, dating Nov 15, 2024.
> >
> >I can act as RM.
> >
> >Best
> >Martin

Re: [DISCUSS] Release pre-trained OpenNLP Models (1.2) ?

Posted by Richard Zowalla <ri...@zowalla.com>.
+1 (do it)

Am 17. November 2024 10:48:18 MEZ schrieb Martin Wiesner <ma...@apache.org>:
>Hi,
>
>Any objections against (or support for) doing a release (1.2) for the pre-trained models? Details below:
>
>I’ve added 9 new languages (via UD treebanks) to our model list:
>
>- "Armenian|hy|BSUT"
>- "Basque|eu|BDT"
>- "Catalan|ca|AnCora"
>- "Georgian|ka|GLC"
>- "Greek|el|GDT"
>- "Kazakh|kk|KTB"
>- "Korean|ko|Kaist"
>- "Icelandic|is|IcePaHC" 
>- "Turkish|tr|BOUN"
>
>In total, this results in 32 supported languages, each with sentence detection, tokenization and POS tagging.
>
>Moreover, the updated ud-train script now produces ME models for the Lemmatizer component, 32 of them.
>
>The training is conducted with OpenNLP 2.5.0 and with the treebanks from the latest UD release, dating Nov 15, 2024.
>
>I can act as RM.
>
>Best
>Martin