From a26f864ed3c227fab1d2a506e27cb4b5b5d831d2 Mon Sep 17 00:00:00 2001 From: Marek Grzenkowicz Date: Tue, 8 Sep 2020 21:13:50 +0200 Subject: [PATCH] Clarify how to choose pretrained weights files (closes #6027) [ci skip] (#6039) --- website/docs/api/cli.md | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/website/docs/api/cli.md b/website/docs/api/cli.md index 779fa7695..b97308aab 100644 --- a/website/docs/api/cli.md +++ b/website/docs/api/cli.md @@ -445,7 +445,8 @@ an approximate language-modeling objective. Specifically, we load pretrained vectors, and train a component like a CNN, BiLSTM, etc to predict vectors which match the pretrained ones. The weights are saved to a directory after each epoch. You can then pass a path to one of these pretrained weights files to the -`spacy train` command. +`spacy train` command. You can try to use a few with low `Loss` values reported +in the output. This technique may be especially helpful if you have little labelled data. However, it's still quite experimental, so your mileage may vary. To load the