explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-04-21 09:31:59 +03:00
Code Issues Packages Projects Releases Wiki Activity

Add clarification to dev docs

Browse Source
This commit is contained in:
shadeMe 2023-07-04 17:21:56 +02:00
parent 725b83bba7
commit 47234f58ed
No known key found for this signature in database
GPG Key ID: 6FCA9FC635B2A402
1 changed files with 9 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
extra/DEVELOPER_DOCS/Listeners.md
View File

@ -223,9 +223,15 @@ The new config and model are then properly stored on the `nlp` object.
Note that this functionality (running the replacement for a transformer listener) was broken prior to
`spacy-transformers` 1.0.5.
As of spaCy 3.7, the `replace_listener` callback accepts three arguments instead of just one: the new (copied) model,
the listener to be replaced, the `tok2vec`/`transformer` pipe from which the new model was copied.
In spaCy 3.7, `Language.replace_listeners` was updated to pass the following additional arguments to the `replace_listener` callback:
the listener to be replaced, the `tok2vec`/`transformer` pipe from which the new model was copied. To maintain backwards-compatiblity,
the method only passes these extra arguments for callbacks that support them:
```
new_model = tok2vec_model.attrs["replace_listener"](new_model, replaced_listener, tok2vec)
def replace_listener_pre_37(copied_tok2vec_model):
...
def replace_listener_post_37(copied_tok2vec_model, replaced_listener, tok2vec_pipe):
...
```