explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-04-21 17:41:59 +03:00
Code Issues Packages Projects Releases Wiki Activity

Add to docs

Browse Source
This commit is contained in:
thomashacker 2023-03-31 10:56:49 +02:00
parent bce15df01c
commit be24c0a0b7
1 changed files with 7 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
website/docs/usage/embeddings-transformers.mdx
View File

@ -745,14 +745,17 @@ To benefit from pretraining, your training step needs to know to initialize its
this by setting `initialize.init_tok2vec` to the filename of the `.bin` file
that you want to use from pretraining.
A pretraining step that runs for 5 epochs with an output path of `pretrain/`, as
an example, produces `pretrain/model0.bin` through `pretrain/model4.bin`. To
make use of the final output, you could fill in this value in your config file:
Similar to training, pretraining produces a `model-last.bin` file which is the
last iteration of the trained weights which you can use to initialize your
`tok2vec` layer. Additionally, you can configure `n_save_epoch` to tell
pretraining in which epoch interval it should save the current training
progress. To make use of the final output, you could fill in this value in your
config file:
```ini {title="config.cfg"}
[paths]
init_tok2vec = "pretrain/model4.bin"
init_tok2vec = "pretrain/model-last.bin"
[initialize]
init_tok2vec = ${paths.init_tok2vec}