Update model card for 3.9.6
Browse files
README.md
CHANGED
|
@@ -80,7 +80,7 @@ model-index:
|
|
| 80 |
| Feature | Description |
|
| 81 |
| --- | --- |
|
| 82 |
| **Name** | `la_core_web_sm` |
|
| 83 |
-
| **Version** | `3.9.
|
| 84 |
| **spaCy** | `>=3.8.11,<3.9.0` |
|
| 85 |
| **Default Pipeline** | `enclitic_splitter`, `tok2vec`, `senter`, `token_fix`, `normer`, `tagger`, `morphologizer`, `trainable_lemmatizer`, `lookup_lemmatizer`, `uv_normalizer`, `parser`, `harmonizer`, `remorpher`, `ner` |
|
| 86 |
| **Components** | `enclitic_splitter`, `tok2vec`, `senter`, `token_fix`, `normer`, `tagger`, `morphologizer`, `trainable_lemmatizer`, `lookup_lemmatizer`, `uv_normalizer`, `parser`, `harmonizer`, `remorpher`, `ner` |
|
|
@@ -90,6 +90,27 @@ model-index:
|
|
| 90 |
| **Author** | [Patrick J. Burns](https://github.com/diyclassics/latincy) |
|
| 91 |
| **Contributors** | Tim Geelhaar (annotation, error analysis [v3.5.2: morphologizer, tagger, parser]); Nora Bernhardt (NER); Vincent Koch (NER) |
|
| 92 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 93 |
## What's new in 3.9.5
|
| 94 |
|
| 95 |
This is a maintenance release across the LatinCy pipeline family
|
|
|
|
| 80 |
| Feature | Description |
|
| 81 |
| --- | --- |
|
| 82 |
| **Name** | `la_core_web_sm` |
|
| 83 |
+
| **Version** | `3.9.6` |
|
| 84 |
| **spaCy** | `>=3.8.11,<3.9.0` |
|
| 85 |
| **Default Pipeline** | `enclitic_splitter`, `tok2vec`, `senter`, `token_fix`, `normer`, `tagger`, `morphologizer`, `trainable_lemmatizer`, `lookup_lemmatizer`, `uv_normalizer`, `parser`, `harmonizer`, `remorpher`, `ner` |
|
| 86 |
| **Components** | `enclitic_splitter`, `tok2vec`, `senter`, `token_fix`, `normer`, `tagger`, `morphologizer`, `trainable_lemmatizer`, `lookup_lemmatizer`, `uv_normalizer`, `parser`, `harmonizer`, `remorpher`, `ner` |
|
|
|
|
| 90 |
| **Author** | [Patrick J. Burns](https://github.com/diyclassics/latincy) |
|
| 91 |
| **Contributors** | Tim Geelhaar (annotation, error analysis [v3.5.2: morphologizer, tagger, parser]); Nora Bernhardt (NER); Vincent Koch (NER) |
|
| 92 |
|
| 93 |
+
## What's new in 3.9.6
|
| 94 |
+
|
| 95 |
+
This is a maintenance release across the LatinCy pipeline family
|
| 96 |
+
(`la_core_web_sm`/`md`/`lg`/`trf`). Model weights are **unchanged** from v3.9.x —
|
| 97 |
+
no retraining; the changes are to sentence segmentation and morphological
|
| 98 |
+
post-processing behavior.
|
| 99 |
+
|
| 100 |
+
- **Fixed sentence segmentation at a terminal followed by a closing single
|
| 101 |
+
quote.** A closing single quote (`'` or `’`) glued directly after a sentence
|
| 102 |
+
terminal (`.`, `?`, `!`, `;`) was suppressing a boundary the senter places for
|
| 103 |
+
the closing *double* quote — so `…non potestis?' Moritur…` was left as a single
|
| 104 |
+
sentence. The `token_fix` corrector now restores the missing boundary, bringing
|
| 105 |
+
`'`/`’` to parity with `"`; direct speech ending in a terminal + closing quote
|
| 106 |
+
is again split from the following clause.
|
| 107 |
+
- **Added a deponent-verb tagging fallback.** When the statistical tagger labels
|
| 108 |
+
a known deponent lemma (e.g. `sequor`, `loquor`, `hortor`) as a plain `verb`,
|
| 109 |
+
the harmonizer now corrects the fine-grained tag to `verb/deponent`, so
|
| 110 |
+
deponents are tagged consistently even when the model misses the subcategory.
|
| 111 |
+
|
| 112 |
+
See [CHANGELOG.md](CHANGELOG.md) for full version history.
|
| 113 |
+
|
| 114 |
## What's new in 3.9.5
|
| 115 |
|
| 116 |
This is a maintenance release across the LatinCy pipeline family
|