SagiPolaczek
commited on
Commit
•
1937fef
1
Parent(s):
75e8c8b
Push model using huggingface_hub.
Browse files
tokenizer/bpe_tokenizer_trained_on_chembl_zinc_with_aug_4272372_samples_balanced_1_1.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer/cell_attributes_tokenizer.json
ADDED
@@ -0,0 +1,3988 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"version": "1.0",
|
3 |
+
"truncation": null,
|
4 |
+
"padding": null,
|
5 |
+
"added_tokens": [
|
6 |
+
{
|
7 |
+
"id": 0,
|
8 |
+
"content": "<UNK>",
|
9 |
+
"single_word": false,
|
10 |
+
"lstrip": false,
|
11 |
+
"rstrip": false,
|
12 |
+
"normalized": false,
|
13 |
+
"special": true
|
14 |
+
},
|
15 |
+
{
|
16 |
+
"id": 1,
|
17 |
+
"content": "<PAD>",
|
18 |
+
"single_word": false,
|
19 |
+
"lstrip": false,
|
20 |
+
"rstrip": false,
|
21 |
+
"normalized": false,
|
22 |
+
"special": true
|
23 |
+
},
|
24 |
+
{
|
25 |
+
"id": 2,
|
26 |
+
"content": "<CLS>",
|
27 |
+
"single_word": false,
|
28 |
+
"lstrip": false,
|
29 |
+
"rstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"special": true
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"id": 3,
|
35 |
+
"content": "<SEP>",
|
36 |
+
"single_word": false,
|
37 |
+
"lstrip": false,
|
38 |
+
"rstrip": false,
|
39 |
+
"normalized": false,
|
40 |
+
"special": true
|
41 |
+
},
|
42 |
+
{
|
43 |
+
"id": 4,
|
44 |
+
"content": "<MASK>",
|
45 |
+
"single_word": false,
|
46 |
+
"lstrip": false,
|
47 |
+
"rstrip": false,
|
48 |
+
"normalized": false,
|
49 |
+
"special": true
|
50 |
+
},
|
51 |
+
{
|
52 |
+
"id": 5,
|
53 |
+
"content": "<EOS>",
|
54 |
+
"single_word": false,
|
55 |
+
"lstrip": false,
|
56 |
+
"rstrip": false,
|
57 |
+
"normalized": false,
|
58 |
+
"special": true
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"id": 6,
|
62 |
+
"content": "<MOLECULAR_ENTITY>",
|
63 |
+
"single_word": false,
|
64 |
+
"lstrip": false,
|
65 |
+
"rstrip": false,
|
66 |
+
"normalized": false,
|
67 |
+
"special": true
|
68 |
+
},
|
69 |
+
{
|
70 |
+
"id": 7,
|
71 |
+
"content": "<GLOBAL_INTERACTION_ATTRIBUTES>",
|
72 |
+
"single_word": false,
|
73 |
+
"lstrip": false,
|
74 |
+
"rstrip": false,
|
75 |
+
"normalized": false,
|
76 |
+
"special": true
|
77 |
+
},
|
78 |
+
{
|
79 |
+
"id": 8,
|
80 |
+
"content": "<MOLECULAR_ENTITY_ANTIGEN>",
|
81 |
+
"single_word": false,
|
82 |
+
"lstrip": false,
|
83 |
+
"rstrip": false,
|
84 |
+
"normalized": false,
|
85 |
+
"special": true
|
86 |
+
},
|
87 |
+
{
|
88 |
+
"id": 9,
|
89 |
+
"content": "<MOLECULAR_ENTITY_EPITOPE>",
|
90 |
+
"single_word": false,
|
91 |
+
"lstrip": false,
|
92 |
+
"rstrip": false,
|
93 |
+
"normalized": false,
|
94 |
+
"special": true
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"id": 10,
|
98 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN>",
|
99 |
+
"single_word": false,
|
100 |
+
"lstrip": false,
|
101 |
+
"rstrip": false,
|
102 |
+
"normalized": false,
|
103 |
+
"special": true
|
104 |
+
},
|
105 |
+
{
|
106 |
+
"id": 11,
|
107 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN>",
|
108 |
+
"single_word": false,
|
109 |
+
"lstrip": false,
|
110 |
+
"rstrip": false,
|
111 |
+
"normalized": false,
|
112 |
+
"special": true
|
113 |
+
},
|
114 |
+
{
|
115 |
+
"id": 12,
|
116 |
+
"content": "<MOLECULAR_ENTITY_TCR_ALPHA_CHAIN>",
|
117 |
+
"single_word": false,
|
118 |
+
"lstrip": false,
|
119 |
+
"rstrip": false,
|
120 |
+
"normalized": false,
|
121 |
+
"special": true
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"id": 13,
|
125 |
+
"content": "<MOLECULAR_ENTITY_TCR_BETA_VDJ>",
|
126 |
+
"single_word": false,
|
127 |
+
"lstrip": false,
|
128 |
+
"rstrip": false,
|
129 |
+
"normalized": false,
|
130 |
+
"special": true
|
131 |
+
},
|
132 |
+
{
|
133 |
+
"id": 14,
|
134 |
+
"content": "<MOLECULAR_ENTITY_TCR_BETA_CDR3>",
|
135 |
+
"single_word": false,
|
136 |
+
"lstrip": false,
|
137 |
+
"rstrip": false,
|
138 |
+
"normalized": false,
|
139 |
+
"special": true
|
140 |
+
},
|
141 |
+
{
|
142 |
+
"id": 15,
|
143 |
+
"content": "<BINDING_AFFINITY_CLASS>",
|
144 |
+
"single_word": false,
|
145 |
+
"lstrip": false,
|
146 |
+
"rstrip": false,
|
147 |
+
"normalized": false,
|
148 |
+
"special": true
|
149 |
+
},
|
150 |
+
{
|
151 |
+
"id": 16,
|
152 |
+
"content": "<DECODER_START>",
|
153 |
+
"single_word": false,
|
154 |
+
"lstrip": false,
|
155 |
+
"rstrip": false,
|
156 |
+
"normalized": false,
|
157 |
+
"special": true
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"id": 17,
|
161 |
+
"content": "<BINDING>",
|
162 |
+
"single_word": false,
|
163 |
+
"lstrip": false,
|
164 |
+
"rstrip": false,
|
165 |
+
"normalized": false,
|
166 |
+
"special": true
|
167 |
+
},
|
168 |
+
{
|
169 |
+
"id": 18,
|
170 |
+
"content": "<FILLIN>",
|
171 |
+
"single_word": false,
|
172 |
+
"lstrip": false,
|
173 |
+
"rstrip": false,
|
174 |
+
"normalized": false,
|
175 |
+
"special": true
|
176 |
+
},
|
177 |
+
{
|
178 |
+
"id": 19,
|
179 |
+
"content": "<REORDER>",
|
180 |
+
"single_word": false,
|
181 |
+
"lstrip": false,
|
182 |
+
"rstrip": false,
|
183 |
+
"normalized": false,
|
184 |
+
"special": true
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"id": 20,
|
188 |
+
"content": "<TOAA>",
|
189 |
+
"single_word": false,
|
190 |
+
"lstrip": false,
|
191 |
+
"rstrip": false,
|
192 |
+
"normalized": false,
|
193 |
+
"special": true
|
194 |
+
},
|
195 |
+
{
|
196 |
+
"id": 21,
|
197 |
+
"content": "<ACTIVE>",
|
198 |
+
"single_word": false,
|
199 |
+
"lstrip": false,
|
200 |
+
"rstrip": false,
|
201 |
+
"normalized": false,
|
202 |
+
"special": true
|
203 |
+
},
|
204 |
+
{
|
205 |
+
"id": 22,
|
206 |
+
"content": "<GENESEQ>",
|
207 |
+
"single_word": false,
|
208 |
+
"lstrip": false,
|
209 |
+
"rstrip": false,
|
210 |
+
"normalized": false,
|
211 |
+
"special": true
|
212 |
+
},
|
213 |
+
{
|
214 |
+
"id": 23,
|
215 |
+
"content": "<INCREASE>",
|
216 |
+
"single_word": false,
|
217 |
+
"lstrip": false,
|
218 |
+
"rstrip": false,
|
219 |
+
"normalized": false,
|
220 |
+
"special": true
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"id": 24,
|
224 |
+
"content": "<DECREASE>",
|
225 |
+
"single_word": false,
|
226 |
+
"lstrip": false,
|
227 |
+
"rstrip": false,
|
228 |
+
"normalized": false,
|
229 |
+
"special": true
|
230 |
+
},
|
231 |
+
{
|
232 |
+
"id": 25,
|
233 |
+
"content": "<STRUCTURE>",
|
234 |
+
"single_word": false,
|
235 |
+
"lstrip": false,
|
236 |
+
"rstrip": false,
|
237 |
+
"normalized": false,
|
238 |
+
"special": true
|
239 |
+
},
|
240 |
+
{
|
241 |
+
"id": 26,
|
242 |
+
"content": "<DISTANCE>",
|
243 |
+
"single_word": false,
|
244 |
+
"lstrip": false,
|
245 |
+
"rstrip": false,
|
246 |
+
"normalized": false,
|
247 |
+
"special": true
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"id": 27,
|
251 |
+
"content": "<SOLUBILITY>",
|
252 |
+
"single_word": false,
|
253 |
+
"lstrip": false,
|
254 |
+
"rstrip": false,
|
255 |
+
"normalized": false,
|
256 |
+
"special": true
|
257 |
+
},
|
258 |
+
{
|
259 |
+
"id": 28,
|
260 |
+
"content": "<TOXICITY>",
|
261 |
+
"single_word": false,
|
262 |
+
"lstrip": false,
|
263 |
+
"rstrip": false,
|
264 |
+
"normalized": false,
|
265 |
+
"special": true
|
266 |
+
},
|
267 |
+
{
|
268 |
+
"id": 29,
|
269 |
+
"content": "<AB>",
|
270 |
+
"single_word": false,
|
271 |
+
"lstrip": false,
|
272 |
+
"rstrip": false,
|
273 |
+
"normalized": false,
|
274 |
+
"special": true
|
275 |
+
},
|
276 |
+
{
|
277 |
+
"id": 30,
|
278 |
+
"content": "<ISACTIVE>",
|
279 |
+
"single_word": false,
|
280 |
+
"lstrip": false,
|
281 |
+
"rstrip": false,
|
282 |
+
"normalized": false,
|
283 |
+
"special": true
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"id": 31,
|
287 |
+
"content": "<ISSYNTHETIC>",
|
288 |
+
"single_word": false,
|
289 |
+
"lstrip": false,
|
290 |
+
"rstrip": false,
|
291 |
+
"normalized": false,
|
292 |
+
"special": true
|
293 |
+
},
|
294 |
+
{
|
295 |
+
"id": 32,
|
296 |
+
"content": "<PENETR>",
|
297 |
+
"single_word": false,
|
298 |
+
"lstrip": false,
|
299 |
+
"rstrip": false,
|
300 |
+
"normalized": false,
|
301 |
+
"special": true
|
302 |
+
},
|
303 |
+
{
|
304 |
+
"id": 33,
|
305 |
+
"content": "<ABSORPTION>",
|
306 |
+
"single_word": false,
|
307 |
+
"lstrip": false,
|
308 |
+
"rstrip": false,
|
309 |
+
"normalized": false,
|
310 |
+
"special": true
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"id": 34,
|
314 |
+
"content": "<DISTRIBUTION>",
|
315 |
+
"single_word": false,
|
316 |
+
"lstrip": false,
|
317 |
+
"rstrip": false,
|
318 |
+
"normalized": false,
|
319 |
+
"special": true
|
320 |
+
},
|
321 |
+
{
|
322 |
+
"id": 35,
|
323 |
+
"content": "<METABOLISM>",
|
324 |
+
"single_word": false,
|
325 |
+
"lstrip": false,
|
326 |
+
"rstrip": false,
|
327 |
+
"normalized": false,
|
328 |
+
"special": true
|
329 |
+
},
|
330 |
+
{
|
331 |
+
"id": 36,
|
332 |
+
"content": "<EXCRETION>",
|
333 |
+
"single_word": false,
|
334 |
+
"lstrip": false,
|
335 |
+
"rstrip": false,
|
336 |
+
"normalized": false,
|
337 |
+
"special": true
|
338 |
+
},
|
339 |
+
{
|
340 |
+
"id": 37,
|
341 |
+
"content": "<FLUORESCENCE>",
|
342 |
+
"single_word": false,
|
343 |
+
"lstrip": false,
|
344 |
+
"rstrip": false,
|
345 |
+
"normalized": false,
|
346 |
+
"special": true
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"id": 38,
|
350 |
+
"content": "<STABILITY>",
|
351 |
+
"single_word": false,
|
352 |
+
"lstrip": false,
|
353 |
+
"rstrip": false,
|
354 |
+
"normalized": false,
|
355 |
+
"special": true
|
356 |
+
},
|
357 |
+
{
|
358 |
+
"id": 39,
|
359 |
+
"content": "<DISORDER>",
|
360 |
+
"single_word": false,
|
361 |
+
"lstrip": false,
|
362 |
+
"rstrip": false,
|
363 |
+
"normalized": false,
|
364 |
+
"special": true
|
365 |
+
},
|
366 |
+
{
|
367 |
+
"id": 40,
|
368 |
+
"content": "<DISEASE>",
|
369 |
+
"single_word": false,
|
370 |
+
"lstrip": false,
|
371 |
+
"rstrip": false,
|
372 |
+
"normalized": false,
|
373 |
+
"special": true
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"id": 41,
|
377 |
+
"content": "<BINARY>",
|
378 |
+
"single_word": false,
|
379 |
+
"lstrip": false,
|
380 |
+
"rstrip": false,
|
381 |
+
"normalized": false,
|
382 |
+
"special": true
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"id": 42,
|
386 |
+
"content": "<REGRESSION>",
|
387 |
+
"single_word": false,
|
388 |
+
"lstrip": false,
|
389 |
+
"rstrip": false,
|
390 |
+
"normalized": false,
|
391 |
+
"special": true
|
392 |
+
},
|
393 |
+
{
|
394 |
+
"id": 43,
|
395 |
+
"content": "<ORGANISM>",
|
396 |
+
"single_word": false,
|
397 |
+
"lstrip": false,
|
398 |
+
"rstrip": false,
|
399 |
+
"normalized": false,
|
400 |
+
"special": true
|
401 |
+
},
|
402 |
+
{
|
403 |
+
"id": 44,
|
404 |
+
"content": "<0>",
|
405 |
+
"single_word": false,
|
406 |
+
"lstrip": false,
|
407 |
+
"rstrip": false,
|
408 |
+
"normalized": false,
|
409 |
+
"special": true
|
410 |
+
},
|
411 |
+
{
|
412 |
+
"id": 45,
|
413 |
+
"content": "<1>",
|
414 |
+
"single_word": false,
|
415 |
+
"lstrip": false,
|
416 |
+
"rstrip": false,
|
417 |
+
"normalized": false,
|
418 |
+
"special": true
|
419 |
+
},
|
420 |
+
{
|
421 |
+
"id": 46,
|
422 |
+
"content": "<2>",
|
423 |
+
"single_word": false,
|
424 |
+
"lstrip": false,
|
425 |
+
"rstrip": false,
|
426 |
+
"normalized": false,
|
427 |
+
"special": true
|
428 |
+
},
|
429 |
+
{
|
430 |
+
"id": 47,
|
431 |
+
"content": "<3>",
|
432 |
+
"single_word": false,
|
433 |
+
"lstrip": false,
|
434 |
+
"rstrip": false,
|
435 |
+
"normalized": false,
|
436 |
+
"special": true
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"id": 48,
|
440 |
+
"content": "<4>",
|
441 |
+
"single_word": false,
|
442 |
+
"lstrip": false,
|
443 |
+
"rstrip": false,
|
444 |
+
"normalized": false,
|
445 |
+
"special": true
|
446 |
+
},
|
447 |
+
{
|
448 |
+
"id": 49,
|
449 |
+
"content": "<5>",
|
450 |
+
"single_word": false,
|
451 |
+
"lstrip": false,
|
452 |
+
"rstrip": false,
|
453 |
+
"normalized": false,
|
454 |
+
"special": true
|
455 |
+
},
|
456 |
+
{
|
457 |
+
"id": 50,
|
458 |
+
"content": "<6>",
|
459 |
+
"single_word": false,
|
460 |
+
"lstrip": false,
|
461 |
+
"rstrip": false,
|
462 |
+
"normalized": false,
|
463 |
+
"special": true
|
464 |
+
},
|
465 |
+
{
|
466 |
+
"id": 51,
|
467 |
+
"content": "<7>",
|
468 |
+
"single_word": false,
|
469 |
+
"lstrip": false,
|
470 |
+
"rstrip": false,
|
471 |
+
"normalized": false,
|
472 |
+
"special": true
|
473 |
+
},
|
474 |
+
{
|
475 |
+
"id": 52,
|
476 |
+
"content": "<8>",
|
477 |
+
"single_word": false,
|
478 |
+
"lstrip": false,
|
479 |
+
"rstrip": false,
|
480 |
+
"normalized": false,
|
481 |
+
"special": true
|
482 |
+
},
|
483 |
+
{
|
484 |
+
"id": 53,
|
485 |
+
"content": "<9>",
|
486 |
+
"single_word": false,
|
487 |
+
"lstrip": false,
|
488 |
+
"rstrip": false,
|
489 |
+
"normalized": false,
|
490 |
+
"special": true
|
491 |
+
},
|
492 |
+
{
|
493 |
+
"id": 54,
|
494 |
+
"content": "<.>",
|
495 |
+
"single_word": false,
|
496 |
+
"lstrip": false,
|
497 |
+
"rstrip": false,
|
498 |
+
"normalized": false,
|
499 |
+
"special": true
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"id": 55,
|
503 |
+
"content": "<YES>",
|
504 |
+
"single_word": false,
|
505 |
+
"lstrip": false,
|
506 |
+
"rstrip": false,
|
507 |
+
"normalized": false,
|
508 |
+
"special": true
|
509 |
+
},
|
510 |
+
{
|
511 |
+
"id": 56,
|
512 |
+
"content": "<NO>",
|
513 |
+
"single_word": false,
|
514 |
+
"lstrip": false,
|
515 |
+
"rstrip": false,
|
516 |
+
"normalized": false,
|
517 |
+
"special": true
|
518 |
+
},
|
519 |
+
{
|
520 |
+
"id": 57,
|
521 |
+
"content": "<SENTINEL_ID_0>",
|
522 |
+
"single_word": false,
|
523 |
+
"lstrip": false,
|
524 |
+
"rstrip": false,
|
525 |
+
"normalized": false,
|
526 |
+
"special": true
|
527 |
+
},
|
528 |
+
{
|
529 |
+
"id": 58,
|
530 |
+
"content": "<SENTINEL_ID_1>",
|
531 |
+
"single_word": false,
|
532 |
+
"lstrip": false,
|
533 |
+
"rstrip": false,
|
534 |
+
"normalized": false,
|
535 |
+
"special": true
|
536 |
+
},
|
537 |
+
{
|
538 |
+
"id": 59,
|
539 |
+
"content": "<SENTINEL_ID_2>",
|
540 |
+
"single_word": false,
|
541 |
+
"lstrip": false,
|
542 |
+
"rstrip": false,
|
543 |
+
"normalized": false,
|
544 |
+
"special": true
|
545 |
+
},
|
546 |
+
{
|
547 |
+
"id": 60,
|
548 |
+
"content": "<SENTINEL_ID_3>",
|
549 |
+
"single_word": false,
|
550 |
+
"lstrip": false,
|
551 |
+
"rstrip": false,
|
552 |
+
"normalized": false,
|
553 |
+
"special": true
|
554 |
+
},
|
555 |
+
{
|
556 |
+
"id": 61,
|
557 |
+
"content": "<SENTINEL_ID_4>",
|
558 |
+
"single_word": false,
|
559 |
+
"lstrip": false,
|
560 |
+
"rstrip": false,
|
561 |
+
"normalized": false,
|
562 |
+
"special": true
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"id": 62,
|
566 |
+
"content": "<SENTINEL_ID_5>",
|
567 |
+
"single_word": false,
|
568 |
+
"lstrip": false,
|
569 |
+
"rstrip": false,
|
570 |
+
"normalized": false,
|
571 |
+
"special": true
|
572 |
+
},
|
573 |
+
{
|
574 |
+
"id": 63,
|
575 |
+
"content": "<SENTINEL_ID_6>",
|
576 |
+
"single_word": false,
|
577 |
+
"lstrip": false,
|
578 |
+
"rstrip": false,
|
579 |
+
"normalized": false,
|
580 |
+
"special": true
|
581 |
+
},
|
582 |
+
{
|
583 |
+
"id": 64,
|
584 |
+
"content": "<SENTINEL_ID_7>",
|
585 |
+
"single_word": false,
|
586 |
+
"lstrip": false,
|
587 |
+
"rstrip": false,
|
588 |
+
"normalized": false,
|
589 |
+
"special": true
|
590 |
+
},
|
591 |
+
{
|
592 |
+
"id": 65,
|
593 |
+
"content": "<SENTINEL_ID_8>",
|
594 |
+
"single_word": false,
|
595 |
+
"lstrip": false,
|
596 |
+
"rstrip": false,
|
597 |
+
"normalized": false,
|
598 |
+
"special": true
|
599 |
+
},
|
600 |
+
{
|
601 |
+
"id": 66,
|
602 |
+
"content": "<SENTINEL_ID_9>",
|
603 |
+
"single_word": false,
|
604 |
+
"lstrip": false,
|
605 |
+
"rstrip": false,
|
606 |
+
"normalized": false,
|
607 |
+
"special": true
|
608 |
+
},
|
609 |
+
{
|
610 |
+
"id": 67,
|
611 |
+
"content": "<SENTINEL_ID_10>",
|
612 |
+
"single_word": false,
|
613 |
+
"lstrip": false,
|
614 |
+
"rstrip": false,
|
615 |
+
"normalized": false,
|
616 |
+
"special": true
|
617 |
+
},
|
618 |
+
{
|
619 |
+
"id": 68,
|
620 |
+
"content": "<SENTINEL_ID_11>",
|
621 |
+
"single_word": false,
|
622 |
+
"lstrip": false,
|
623 |
+
"rstrip": false,
|
624 |
+
"normalized": false,
|
625 |
+
"special": true
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"id": 69,
|
629 |
+
"content": "<SENTINEL_ID_12>",
|
630 |
+
"single_word": false,
|
631 |
+
"lstrip": false,
|
632 |
+
"rstrip": false,
|
633 |
+
"normalized": false,
|
634 |
+
"special": true
|
635 |
+
},
|
636 |
+
{
|
637 |
+
"id": 70,
|
638 |
+
"content": "<SENTINEL_ID_13>",
|
639 |
+
"single_word": false,
|
640 |
+
"lstrip": false,
|
641 |
+
"rstrip": false,
|
642 |
+
"normalized": false,
|
643 |
+
"special": true
|
644 |
+
},
|
645 |
+
{
|
646 |
+
"id": 71,
|
647 |
+
"content": "<SENTINEL_ID_14>",
|
648 |
+
"single_word": false,
|
649 |
+
"lstrip": false,
|
650 |
+
"rstrip": false,
|
651 |
+
"normalized": false,
|
652 |
+
"special": true
|
653 |
+
},
|
654 |
+
{
|
655 |
+
"id": 72,
|
656 |
+
"content": "<SENTINEL_ID_15>",
|
657 |
+
"single_word": false,
|
658 |
+
"lstrip": false,
|
659 |
+
"rstrip": false,
|
660 |
+
"normalized": false,
|
661 |
+
"special": true
|
662 |
+
},
|
663 |
+
{
|
664 |
+
"id": 73,
|
665 |
+
"content": "<SENTINEL_ID_16>",
|
666 |
+
"single_word": false,
|
667 |
+
"lstrip": false,
|
668 |
+
"rstrip": false,
|
669 |
+
"normalized": false,
|
670 |
+
"special": true
|
671 |
+
},
|
672 |
+
{
|
673 |
+
"id": 74,
|
674 |
+
"content": "<SENTINEL_ID_17>",
|
675 |
+
"single_word": false,
|
676 |
+
"lstrip": false,
|
677 |
+
"rstrip": false,
|
678 |
+
"normalized": false,
|
679 |
+
"special": true
|
680 |
+
},
|
681 |
+
{
|
682 |
+
"id": 75,
|
683 |
+
"content": "<SENTINEL_ID_18>",
|
684 |
+
"single_word": false,
|
685 |
+
"lstrip": false,
|
686 |
+
"rstrip": false,
|
687 |
+
"normalized": false,
|
688 |
+
"special": true
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"id": 76,
|
692 |
+
"content": "<SENTINEL_ID_19>",
|
693 |
+
"single_word": false,
|
694 |
+
"lstrip": false,
|
695 |
+
"rstrip": false,
|
696 |
+
"normalized": false,
|
697 |
+
"special": true
|
698 |
+
},
|
699 |
+
{
|
700 |
+
"id": 77,
|
701 |
+
"content": "<SENTINEL_ID_20>",
|
702 |
+
"single_word": false,
|
703 |
+
"lstrip": false,
|
704 |
+
"rstrip": false,
|
705 |
+
"normalized": false,
|
706 |
+
"special": true
|
707 |
+
},
|
708 |
+
{
|
709 |
+
"id": 78,
|
710 |
+
"content": "<SENTINEL_ID_21>",
|
711 |
+
"single_word": false,
|
712 |
+
"lstrip": false,
|
713 |
+
"rstrip": false,
|
714 |
+
"normalized": false,
|
715 |
+
"special": true
|
716 |
+
},
|
717 |
+
{
|
718 |
+
"id": 79,
|
719 |
+
"content": "<SENTINEL_ID_22>",
|
720 |
+
"single_word": false,
|
721 |
+
"lstrip": false,
|
722 |
+
"rstrip": false,
|
723 |
+
"normalized": false,
|
724 |
+
"special": true
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"id": 80,
|
728 |
+
"content": "<SENTINEL_ID_23>",
|
729 |
+
"single_word": false,
|
730 |
+
"lstrip": false,
|
731 |
+
"rstrip": false,
|
732 |
+
"normalized": false,
|
733 |
+
"special": true
|
734 |
+
},
|
735 |
+
{
|
736 |
+
"id": 81,
|
737 |
+
"content": "<SENTINEL_ID_24>",
|
738 |
+
"single_word": false,
|
739 |
+
"lstrip": false,
|
740 |
+
"rstrip": false,
|
741 |
+
"normalized": false,
|
742 |
+
"special": true
|
743 |
+
},
|
744 |
+
{
|
745 |
+
"id": 82,
|
746 |
+
"content": "<SENTINEL_ID_25>",
|
747 |
+
"single_word": false,
|
748 |
+
"lstrip": false,
|
749 |
+
"rstrip": false,
|
750 |
+
"normalized": false,
|
751 |
+
"special": true
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"id": 83,
|
755 |
+
"content": "<SENTINEL_ID_26>",
|
756 |
+
"single_word": false,
|
757 |
+
"lstrip": false,
|
758 |
+
"rstrip": false,
|
759 |
+
"normalized": false,
|
760 |
+
"special": true
|
761 |
+
},
|
762 |
+
{
|
763 |
+
"id": 84,
|
764 |
+
"content": "<SENTINEL_ID_27>",
|
765 |
+
"single_word": false,
|
766 |
+
"lstrip": false,
|
767 |
+
"rstrip": false,
|
768 |
+
"normalized": false,
|
769 |
+
"special": true
|
770 |
+
},
|
771 |
+
{
|
772 |
+
"id": 85,
|
773 |
+
"content": "<SENTINEL_ID_28>",
|
774 |
+
"single_word": false,
|
775 |
+
"lstrip": false,
|
776 |
+
"rstrip": false,
|
777 |
+
"normalized": false,
|
778 |
+
"special": true
|
779 |
+
},
|
780 |
+
{
|
781 |
+
"id": 86,
|
782 |
+
"content": "<SENTINEL_ID_29>",
|
783 |
+
"single_word": false,
|
784 |
+
"lstrip": false,
|
785 |
+
"rstrip": false,
|
786 |
+
"normalized": false,
|
787 |
+
"special": true
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"id": 87,
|
791 |
+
"content": "<SENTINEL_ID_30>",
|
792 |
+
"single_word": false,
|
793 |
+
"lstrip": false,
|
794 |
+
"rstrip": false,
|
795 |
+
"normalized": false,
|
796 |
+
"special": true
|
797 |
+
},
|
798 |
+
{
|
799 |
+
"id": 88,
|
800 |
+
"content": "<SENTINEL_ID_31>",
|
801 |
+
"single_word": false,
|
802 |
+
"lstrip": false,
|
803 |
+
"rstrip": false,
|
804 |
+
"normalized": false,
|
805 |
+
"special": true
|
806 |
+
},
|
807 |
+
{
|
808 |
+
"id": 89,
|
809 |
+
"content": "<SENTINEL_ID_32>",
|
810 |
+
"single_word": false,
|
811 |
+
"lstrip": false,
|
812 |
+
"rstrip": false,
|
813 |
+
"normalized": false,
|
814 |
+
"special": true
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"id": 90,
|
818 |
+
"content": "<SENTINEL_ID_33>",
|
819 |
+
"single_word": false,
|
820 |
+
"lstrip": false,
|
821 |
+
"rstrip": false,
|
822 |
+
"normalized": false,
|
823 |
+
"special": true
|
824 |
+
},
|
825 |
+
{
|
826 |
+
"id": 91,
|
827 |
+
"content": "<SENTINEL_ID_34>",
|
828 |
+
"single_word": false,
|
829 |
+
"lstrip": false,
|
830 |
+
"rstrip": false,
|
831 |
+
"normalized": false,
|
832 |
+
"special": true
|
833 |
+
},
|
834 |
+
{
|
835 |
+
"id": 92,
|
836 |
+
"content": "<SENTINEL_ID_35>",
|
837 |
+
"single_word": false,
|
838 |
+
"lstrip": false,
|
839 |
+
"rstrip": false,
|
840 |
+
"normalized": false,
|
841 |
+
"special": true
|
842 |
+
},
|
843 |
+
{
|
844 |
+
"id": 93,
|
845 |
+
"content": "<SENTINEL_ID_36>",
|
846 |
+
"single_word": false,
|
847 |
+
"lstrip": false,
|
848 |
+
"rstrip": false,
|
849 |
+
"normalized": false,
|
850 |
+
"special": true
|
851 |
+
},
|
852 |
+
{
|
853 |
+
"id": 94,
|
854 |
+
"content": "<SENTINEL_ID_37>",
|
855 |
+
"single_word": false,
|
856 |
+
"lstrip": false,
|
857 |
+
"rstrip": false,
|
858 |
+
"normalized": false,
|
859 |
+
"special": true
|
860 |
+
},
|
861 |
+
{
|
862 |
+
"id": 95,
|
863 |
+
"content": "<SENTINEL_ID_38>",
|
864 |
+
"single_word": false,
|
865 |
+
"lstrip": false,
|
866 |
+
"rstrip": false,
|
867 |
+
"normalized": false,
|
868 |
+
"special": true
|
869 |
+
},
|
870 |
+
{
|
871 |
+
"id": 96,
|
872 |
+
"content": "<SENTINEL_ID_39>",
|
873 |
+
"single_word": false,
|
874 |
+
"lstrip": false,
|
875 |
+
"rstrip": false,
|
876 |
+
"normalized": false,
|
877 |
+
"special": true
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"id": 97,
|
881 |
+
"content": "<SENTINEL_ID_40>",
|
882 |
+
"single_word": false,
|
883 |
+
"lstrip": false,
|
884 |
+
"rstrip": false,
|
885 |
+
"normalized": false,
|
886 |
+
"special": true
|
887 |
+
},
|
888 |
+
{
|
889 |
+
"id": 98,
|
890 |
+
"content": "<SENTINEL_ID_41>",
|
891 |
+
"single_word": false,
|
892 |
+
"lstrip": false,
|
893 |
+
"rstrip": false,
|
894 |
+
"normalized": false,
|
895 |
+
"special": true
|
896 |
+
},
|
897 |
+
{
|
898 |
+
"id": 99,
|
899 |
+
"content": "<SENTINEL_ID_42>",
|
900 |
+
"single_word": false,
|
901 |
+
"lstrip": false,
|
902 |
+
"rstrip": false,
|
903 |
+
"normalized": false,
|
904 |
+
"special": true
|
905 |
+
},
|
906 |
+
{
|
907 |
+
"id": 100,
|
908 |
+
"content": "<SENTINEL_ID_43>",
|
909 |
+
"single_word": false,
|
910 |
+
"lstrip": false,
|
911 |
+
"rstrip": false,
|
912 |
+
"normalized": false,
|
913 |
+
"special": true
|
914 |
+
},
|
915 |
+
{
|
916 |
+
"id": 101,
|
917 |
+
"content": "<SENTINEL_ID_44>",
|
918 |
+
"single_word": false,
|
919 |
+
"lstrip": false,
|
920 |
+
"rstrip": false,
|
921 |
+
"normalized": false,
|
922 |
+
"special": true
|
923 |
+
},
|
924 |
+
{
|
925 |
+
"id": 102,
|
926 |
+
"content": "<SENTINEL_ID_45>",
|
927 |
+
"single_word": false,
|
928 |
+
"lstrip": false,
|
929 |
+
"rstrip": false,
|
930 |
+
"normalized": false,
|
931 |
+
"special": true
|
932 |
+
},
|
933 |
+
{
|
934 |
+
"id": 103,
|
935 |
+
"content": "<SENTINEL_ID_46>",
|
936 |
+
"single_word": false,
|
937 |
+
"lstrip": false,
|
938 |
+
"rstrip": false,
|
939 |
+
"normalized": false,
|
940 |
+
"special": true
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"id": 104,
|
944 |
+
"content": "<SENTINEL_ID_47>",
|
945 |
+
"single_word": false,
|
946 |
+
"lstrip": false,
|
947 |
+
"rstrip": false,
|
948 |
+
"normalized": false,
|
949 |
+
"special": true
|
950 |
+
},
|
951 |
+
{
|
952 |
+
"id": 105,
|
953 |
+
"content": "<SENTINEL_ID_48>",
|
954 |
+
"single_word": false,
|
955 |
+
"lstrip": false,
|
956 |
+
"rstrip": false,
|
957 |
+
"normalized": false,
|
958 |
+
"special": true
|
959 |
+
},
|
960 |
+
{
|
961 |
+
"id": 106,
|
962 |
+
"content": "<SENTINEL_ID_49>",
|
963 |
+
"single_word": false,
|
964 |
+
"lstrip": false,
|
965 |
+
"rstrip": false,
|
966 |
+
"normalized": false,
|
967 |
+
"special": true
|
968 |
+
},
|
969 |
+
{
|
970 |
+
"id": 107,
|
971 |
+
"content": "<SENTINEL_ID_50>",
|
972 |
+
"single_word": false,
|
973 |
+
"lstrip": false,
|
974 |
+
"rstrip": false,
|
975 |
+
"normalized": false,
|
976 |
+
"special": true
|
977 |
+
},
|
978 |
+
{
|
979 |
+
"id": 108,
|
980 |
+
"content": "<SENTINEL_ID_51>",
|
981 |
+
"single_word": false,
|
982 |
+
"lstrip": false,
|
983 |
+
"rstrip": false,
|
984 |
+
"normalized": false,
|
985 |
+
"special": true
|
986 |
+
},
|
987 |
+
{
|
988 |
+
"id": 109,
|
989 |
+
"content": "<SENTINEL_ID_52>",
|
990 |
+
"single_word": false,
|
991 |
+
"lstrip": false,
|
992 |
+
"rstrip": false,
|
993 |
+
"normalized": false,
|
994 |
+
"special": true
|
995 |
+
},
|
996 |
+
{
|
997 |
+
"id": 110,
|
998 |
+
"content": "<SENTINEL_ID_53>",
|
999 |
+
"single_word": false,
|
1000 |
+
"lstrip": false,
|
1001 |
+
"rstrip": false,
|
1002 |
+
"normalized": false,
|
1003 |
+
"special": true
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"id": 111,
|
1007 |
+
"content": "<SENTINEL_ID_54>",
|
1008 |
+
"single_word": false,
|
1009 |
+
"lstrip": false,
|
1010 |
+
"rstrip": false,
|
1011 |
+
"normalized": false,
|
1012 |
+
"special": true
|
1013 |
+
},
|
1014 |
+
{
|
1015 |
+
"id": 112,
|
1016 |
+
"content": "<SENTINEL_ID_55>",
|
1017 |
+
"single_word": false,
|
1018 |
+
"lstrip": false,
|
1019 |
+
"rstrip": false,
|
1020 |
+
"normalized": false,
|
1021 |
+
"special": true
|
1022 |
+
},
|
1023 |
+
{
|
1024 |
+
"id": 113,
|
1025 |
+
"content": "<SENTINEL_ID_56>",
|
1026 |
+
"single_word": false,
|
1027 |
+
"lstrip": false,
|
1028 |
+
"rstrip": false,
|
1029 |
+
"normalized": false,
|
1030 |
+
"special": true
|
1031 |
+
},
|
1032 |
+
{
|
1033 |
+
"id": 114,
|
1034 |
+
"content": "<SENTINEL_ID_57>",
|
1035 |
+
"single_word": false,
|
1036 |
+
"lstrip": false,
|
1037 |
+
"rstrip": false,
|
1038 |
+
"normalized": false,
|
1039 |
+
"special": true
|
1040 |
+
},
|
1041 |
+
{
|
1042 |
+
"id": 115,
|
1043 |
+
"content": "<SENTINEL_ID_58>",
|
1044 |
+
"single_word": false,
|
1045 |
+
"lstrip": false,
|
1046 |
+
"rstrip": false,
|
1047 |
+
"normalized": false,
|
1048 |
+
"special": true
|
1049 |
+
},
|
1050 |
+
{
|
1051 |
+
"id": 116,
|
1052 |
+
"content": "<SENTINEL_ID_59>",
|
1053 |
+
"single_word": false,
|
1054 |
+
"lstrip": false,
|
1055 |
+
"rstrip": false,
|
1056 |
+
"normalized": false,
|
1057 |
+
"special": true
|
1058 |
+
},
|
1059 |
+
{
|
1060 |
+
"id": 117,
|
1061 |
+
"content": "<SENTINEL_ID_60>",
|
1062 |
+
"single_word": false,
|
1063 |
+
"lstrip": false,
|
1064 |
+
"rstrip": false,
|
1065 |
+
"normalized": false,
|
1066 |
+
"special": true
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"id": 118,
|
1070 |
+
"content": "<SENTINEL_ID_61>",
|
1071 |
+
"single_word": false,
|
1072 |
+
"lstrip": false,
|
1073 |
+
"rstrip": false,
|
1074 |
+
"normalized": false,
|
1075 |
+
"special": true
|
1076 |
+
},
|
1077 |
+
{
|
1078 |
+
"id": 119,
|
1079 |
+
"content": "<SENTINEL_ID_62>",
|
1080 |
+
"single_word": false,
|
1081 |
+
"lstrip": false,
|
1082 |
+
"rstrip": false,
|
1083 |
+
"normalized": false,
|
1084 |
+
"special": true
|
1085 |
+
},
|
1086 |
+
{
|
1087 |
+
"id": 120,
|
1088 |
+
"content": "<SENTINEL_ID_63>",
|
1089 |
+
"single_word": false,
|
1090 |
+
"lstrip": false,
|
1091 |
+
"rstrip": false,
|
1092 |
+
"normalized": false,
|
1093 |
+
"special": true
|
1094 |
+
},
|
1095 |
+
{
|
1096 |
+
"id": 121,
|
1097 |
+
"content": "<SENTINEL_ID_64>",
|
1098 |
+
"single_word": false,
|
1099 |
+
"lstrip": false,
|
1100 |
+
"rstrip": false,
|
1101 |
+
"normalized": false,
|
1102 |
+
"special": true
|
1103 |
+
},
|
1104 |
+
{
|
1105 |
+
"id": 122,
|
1106 |
+
"content": "<SENTINEL_ID_65>",
|
1107 |
+
"single_word": false,
|
1108 |
+
"lstrip": false,
|
1109 |
+
"rstrip": false,
|
1110 |
+
"normalized": false,
|
1111 |
+
"special": true
|
1112 |
+
},
|
1113 |
+
{
|
1114 |
+
"id": 123,
|
1115 |
+
"content": "<SENTINEL_ID_66>",
|
1116 |
+
"single_word": false,
|
1117 |
+
"lstrip": false,
|
1118 |
+
"rstrip": false,
|
1119 |
+
"normalized": false,
|
1120 |
+
"special": true
|
1121 |
+
},
|
1122 |
+
{
|
1123 |
+
"id": 124,
|
1124 |
+
"content": "<SENTINEL_ID_67>",
|
1125 |
+
"single_word": false,
|
1126 |
+
"lstrip": false,
|
1127 |
+
"rstrip": false,
|
1128 |
+
"normalized": false,
|
1129 |
+
"special": true
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"id": 125,
|
1133 |
+
"content": "<SENTINEL_ID_68>",
|
1134 |
+
"single_word": false,
|
1135 |
+
"lstrip": false,
|
1136 |
+
"rstrip": false,
|
1137 |
+
"normalized": false,
|
1138 |
+
"special": true
|
1139 |
+
},
|
1140 |
+
{
|
1141 |
+
"id": 126,
|
1142 |
+
"content": "<SENTINEL_ID_69>",
|
1143 |
+
"single_word": false,
|
1144 |
+
"lstrip": false,
|
1145 |
+
"rstrip": false,
|
1146 |
+
"normalized": false,
|
1147 |
+
"special": true
|
1148 |
+
},
|
1149 |
+
{
|
1150 |
+
"id": 127,
|
1151 |
+
"content": "<SENTINEL_ID_70>",
|
1152 |
+
"single_word": false,
|
1153 |
+
"lstrip": false,
|
1154 |
+
"rstrip": false,
|
1155 |
+
"normalized": false,
|
1156 |
+
"special": true
|
1157 |
+
},
|
1158 |
+
{
|
1159 |
+
"id": 128,
|
1160 |
+
"content": "<SENTINEL_ID_71>",
|
1161 |
+
"single_word": false,
|
1162 |
+
"lstrip": false,
|
1163 |
+
"rstrip": false,
|
1164 |
+
"normalized": false,
|
1165 |
+
"special": true
|
1166 |
+
},
|
1167 |
+
{
|
1168 |
+
"id": 129,
|
1169 |
+
"content": "<SENTINEL_ID_72>",
|
1170 |
+
"single_word": false,
|
1171 |
+
"lstrip": false,
|
1172 |
+
"rstrip": false,
|
1173 |
+
"normalized": false,
|
1174 |
+
"special": true
|
1175 |
+
},
|
1176 |
+
{
|
1177 |
+
"id": 130,
|
1178 |
+
"content": "<SENTINEL_ID_73>",
|
1179 |
+
"single_word": false,
|
1180 |
+
"lstrip": false,
|
1181 |
+
"rstrip": false,
|
1182 |
+
"normalized": false,
|
1183 |
+
"special": true
|
1184 |
+
},
|
1185 |
+
{
|
1186 |
+
"id": 131,
|
1187 |
+
"content": "<SENTINEL_ID_74>",
|
1188 |
+
"single_word": false,
|
1189 |
+
"lstrip": false,
|
1190 |
+
"rstrip": false,
|
1191 |
+
"normalized": false,
|
1192 |
+
"special": true
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"id": 132,
|
1196 |
+
"content": "<SENTINEL_ID_75>",
|
1197 |
+
"single_word": false,
|
1198 |
+
"lstrip": false,
|
1199 |
+
"rstrip": false,
|
1200 |
+
"normalized": false,
|
1201 |
+
"special": true
|
1202 |
+
},
|
1203 |
+
{
|
1204 |
+
"id": 133,
|
1205 |
+
"content": "<SENTINEL_ID_76>",
|
1206 |
+
"single_word": false,
|
1207 |
+
"lstrip": false,
|
1208 |
+
"rstrip": false,
|
1209 |
+
"normalized": false,
|
1210 |
+
"special": true
|
1211 |
+
},
|
1212 |
+
{
|
1213 |
+
"id": 134,
|
1214 |
+
"content": "<SENTINEL_ID_77>",
|
1215 |
+
"single_word": false,
|
1216 |
+
"lstrip": false,
|
1217 |
+
"rstrip": false,
|
1218 |
+
"normalized": false,
|
1219 |
+
"special": true
|
1220 |
+
},
|
1221 |
+
{
|
1222 |
+
"id": 135,
|
1223 |
+
"content": "<SENTINEL_ID_78>",
|
1224 |
+
"single_word": false,
|
1225 |
+
"lstrip": false,
|
1226 |
+
"rstrip": false,
|
1227 |
+
"normalized": false,
|
1228 |
+
"special": true
|
1229 |
+
},
|
1230 |
+
{
|
1231 |
+
"id": 136,
|
1232 |
+
"content": "<SENTINEL_ID_79>",
|
1233 |
+
"single_word": false,
|
1234 |
+
"lstrip": false,
|
1235 |
+
"rstrip": false,
|
1236 |
+
"normalized": false,
|
1237 |
+
"special": true
|
1238 |
+
},
|
1239 |
+
{
|
1240 |
+
"id": 137,
|
1241 |
+
"content": "<SENTINEL_ID_80>",
|
1242 |
+
"single_word": false,
|
1243 |
+
"lstrip": false,
|
1244 |
+
"rstrip": false,
|
1245 |
+
"normalized": false,
|
1246 |
+
"special": true
|
1247 |
+
},
|
1248 |
+
{
|
1249 |
+
"id": 138,
|
1250 |
+
"content": "<SENTINEL_ID_81>",
|
1251 |
+
"single_word": false,
|
1252 |
+
"lstrip": false,
|
1253 |
+
"rstrip": false,
|
1254 |
+
"normalized": false,
|
1255 |
+
"special": true
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"id": 139,
|
1259 |
+
"content": "<SENTINEL_ID_82>",
|
1260 |
+
"single_word": false,
|
1261 |
+
"lstrip": false,
|
1262 |
+
"rstrip": false,
|
1263 |
+
"normalized": false,
|
1264 |
+
"special": true
|
1265 |
+
},
|
1266 |
+
{
|
1267 |
+
"id": 140,
|
1268 |
+
"content": "<SENTINEL_ID_83>",
|
1269 |
+
"single_word": false,
|
1270 |
+
"lstrip": false,
|
1271 |
+
"rstrip": false,
|
1272 |
+
"normalized": false,
|
1273 |
+
"special": true
|
1274 |
+
},
|
1275 |
+
{
|
1276 |
+
"id": 141,
|
1277 |
+
"content": "<SENTINEL_ID_84>",
|
1278 |
+
"single_word": false,
|
1279 |
+
"lstrip": false,
|
1280 |
+
"rstrip": false,
|
1281 |
+
"normalized": false,
|
1282 |
+
"special": true
|
1283 |
+
},
|
1284 |
+
{
|
1285 |
+
"id": 142,
|
1286 |
+
"content": "<SENTINEL_ID_85>",
|
1287 |
+
"single_word": false,
|
1288 |
+
"lstrip": false,
|
1289 |
+
"rstrip": false,
|
1290 |
+
"normalized": false,
|
1291 |
+
"special": true
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"id": 143,
|
1295 |
+
"content": "<SENTINEL_ID_86>",
|
1296 |
+
"single_word": false,
|
1297 |
+
"lstrip": false,
|
1298 |
+
"rstrip": false,
|
1299 |
+
"normalized": false,
|
1300 |
+
"special": true
|
1301 |
+
},
|
1302 |
+
{
|
1303 |
+
"id": 144,
|
1304 |
+
"content": "<SENTINEL_ID_87>",
|
1305 |
+
"single_word": false,
|
1306 |
+
"lstrip": false,
|
1307 |
+
"rstrip": false,
|
1308 |
+
"normalized": false,
|
1309 |
+
"special": true
|
1310 |
+
},
|
1311 |
+
{
|
1312 |
+
"id": 145,
|
1313 |
+
"content": "<SENTINEL_ID_88>",
|
1314 |
+
"single_word": false,
|
1315 |
+
"lstrip": false,
|
1316 |
+
"rstrip": false,
|
1317 |
+
"normalized": false,
|
1318 |
+
"special": true
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"id": 146,
|
1322 |
+
"content": "<SENTINEL_ID_89>",
|
1323 |
+
"single_word": false,
|
1324 |
+
"lstrip": false,
|
1325 |
+
"rstrip": false,
|
1326 |
+
"normalized": false,
|
1327 |
+
"special": true
|
1328 |
+
},
|
1329 |
+
{
|
1330 |
+
"id": 147,
|
1331 |
+
"content": "<SENTINEL_ID_90>",
|
1332 |
+
"single_word": false,
|
1333 |
+
"lstrip": false,
|
1334 |
+
"rstrip": false,
|
1335 |
+
"normalized": false,
|
1336 |
+
"special": true
|
1337 |
+
},
|
1338 |
+
{
|
1339 |
+
"id": 148,
|
1340 |
+
"content": "<SENTINEL_ID_91>",
|
1341 |
+
"single_word": false,
|
1342 |
+
"lstrip": false,
|
1343 |
+
"rstrip": false,
|
1344 |
+
"normalized": false,
|
1345 |
+
"special": true
|
1346 |
+
},
|
1347 |
+
{
|
1348 |
+
"id": 149,
|
1349 |
+
"content": "<SENTINEL_ID_92>",
|
1350 |
+
"single_word": false,
|
1351 |
+
"lstrip": false,
|
1352 |
+
"rstrip": false,
|
1353 |
+
"normalized": false,
|
1354 |
+
"special": true
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"id": 150,
|
1358 |
+
"content": "<SENTINEL_ID_93>",
|
1359 |
+
"single_word": false,
|
1360 |
+
"lstrip": false,
|
1361 |
+
"rstrip": false,
|
1362 |
+
"normalized": false,
|
1363 |
+
"special": true
|
1364 |
+
},
|
1365 |
+
{
|
1366 |
+
"id": 151,
|
1367 |
+
"content": "<SENTINEL_ID_94>",
|
1368 |
+
"single_word": false,
|
1369 |
+
"lstrip": false,
|
1370 |
+
"rstrip": false,
|
1371 |
+
"normalized": false,
|
1372 |
+
"special": true
|
1373 |
+
},
|
1374 |
+
{
|
1375 |
+
"id": 152,
|
1376 |
+
"content": "<SENTINEL_ID_95>",
|
1377 |
+
"single_word": false,
|
1378 |
+
"lstrip": false,
|
1379 |
+
"rstrip": false,
|
1380 |
+
"normalized": false,
|
1381 |
+
"special": true
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"id": 153,
|
1385 |
+
"content": "<SENTINEL_ID_96>",
|
1386 |
+
"single_word": false,
|
1387 |
+
"lstrip": false,
|
1388 |
+
"rstrip": false,
|
1389 |
+
"normalized": false,
|
1390 |
+
"special": true
|
1391 |
+
},
|
1392 |
+
{
|
1393 |
+
"id": 154,
|
1394 |
+
"content": "<SENTINEL_ID_97>",
|
1395 |
+
"single_word": false,
|
1396 |
+
"lstrip": false,
|
1397 |
+
"rstrip": false,
|
1398 |
+
"normalized": false,
|
1399 |
+
"special": true
|
1400 |
+
},
|
1401 |
+
{
|
1402 |
+
"id": 155,
|
1403 |
+
"content": "<SENTINEL_ID_98>",
|
1404 |
+
"single_word": false,
|
1405 |
+
"lstrip": false,
|
1406 |
+
"rstrip": false,
|
1407 |
+
"normalized": false,
|
1408 |
+
"special": true
|
1409 |
+
},
|
1410 |
+
{
|
1411 |
+
"id": 156,
|
1412 |
+
"content": "<SENTINEL_ID_99>",
|
1413 |
+
"single_word": false,
|
1414 |
+
"lstrip": false,
|
1415 |
+
"rstrip": false,
|
1416 |
+
"normalized": false,
|
1417 |
+
"special": true
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"id": 157,
|
1421 |
+
"content": "<SENTINEL_ID_100>",
|
1422 |
+
"single_word": false,
|
1423 |
+
"lstrip": false,
|
1424 |
+
"rstrip": false,
|
1425 |
+
"normalized": false,
|
1426 |
+
"special": true
|
1427 |
+
},
|
1428 |
+
{
|
1429 |
+
"id": 158,
|
1430 |
+
"content": "<SENTINEL_ID_101>",
|
1431 |
+
"single_word": false,
|
1432 |
+
"lstrip": false,
|
1433 |
+
"rstrip": false,
|
1434 |
+
"normalized": false,
|
1435 |
+
"special": true
|
1436 |
+
},
|
1437 |
+
{
|
1438 |
+
"id": 159,
|
1439 |
+
"content": "<SENTINEL_ID_102>",
|
1440 |
+
"single_word": false,
|
1441 |
+
"lstrip": false,
|
1442 |
+
"rstrip": false,
|
1443 |
+
"normalized": false,
|
1444 |
+
"special": true
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"id": 160,
|
1448 |
+
"content": "<SENTINEL_ID_103>",
|
1449 |
+
"single_word": false,
|
1450 |
+
"lstrip": false,
|
1451 |
+
"rstrip": false,
|
1452 |
+
"normalized": false,
|
1453 |
+
"special": true
|
1454 |
+
},
|
1455 |
+
{
|
1456 |
+
"id": 161,
|
1457 |
+
"content": "<SENTINEL_ID_104>",
|
1458 |
+
"single_word": false,
|
1459 |
+
"lstrip": false,
|
1460 |
+
"rstrip": false,
|
1461 |
+
"normalized": false,
|
1462 |
+
"special": true
|
1463 |
+
},
|
1464 |
+
{
|
1465 |
+
"id": 162,
|
1466 |
+
"content": "<SENTINEL_ID_105>",
|
1467 |
+
"single_word": false,
|
1468 |
+
"lstrip": false,
|
1469 |
+
"rstrip": false,
|
1470 |
+
"normalized": false,
|
1471 |
+
"special": true
|
1472 |
+
},
|
1473 |
+
{
|
1474 |
+
"id": 163,
|
1475 |
+
"content": "<SENTINEL_ID_106>",
|
1476 |
+
"single_word": false,
|
1477 |
+
"lstrip": false,
|
1478 |
+
"rstrip": false,
|
1479 |
+
"normalized": false,
|
1480 |
+
"special": true
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"id": 164,
|
1484 |
+
"content": "<SENTINEL_ID_107>",
|
1485 |
+
"single_word": false,
|
1486 |
+
"lstrip": false,
|
1487 |
+
"rstrip": false,
|
1488 |
+
"normalized": false,
|
1489 |
+
"special": true
|
1490 |
+
},
|
1491 |
+
{
|
1492 |
+
"id": 165,
|
1493 |
+
"content": "<SENTINEL_ID_108>",
|
1494 |
+
"single_word": false,
|
1495 |
+
"lstrip": false,
|
1496 |
+
"rstrip": false,
|
1497 |
+
"normalized": false,
|
1498 |
+
"special": true
|
1499 |
+
},
|
1500 |
+
{
|
1501 |
+
"id": 166,
|
1502 |
+
"content": "<SENTINEL_ID_109>",
|
1503 |
+
"single_word": false,
|
1504 |
+
"lstrip": false,
|
1505 |
+
"rstrip": false,
|
1506 |
+
"normalized": false,
|
1507 |
+
"special": true
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"id": 167,
|
1511 |
+
"content": "<SENTINEL_ID_110>",
|
1512 |
+
"single_word": false,
|
1513 |
+
"lstrip": false,
|
1514 |
+
"rstrip": false,
|
1515 |
+
"normalized": false,
|
1516 |
+
"special": true
|
1517 |
+
},
|
1518 |
+
{
|
1519 |
+
"id": 168,
|
1520 |
+
"content": "<SENTINEL_ID_111>",
|
1521 |
+
"single_word": false,
|
1522 |
+
"lstrip": false,
|
1523 |
+
"rstrip": false,
|
1524 |
+
"normalized": false,
|
1525 |
+
"special": true
|
1526 |
+
},
|
1527 |
+
{
|
1528 |
+
"id": 169,
|
1529 |
+
"content": "<SENTINEL_ID_112>",
|
1530 |
+
"single_word": false,
|
1531 |
+
"lstrip": false,
|
1532 |
+
"rstrip": false,
|
1533 |
+
"normalized": false,
|
1534 |
+
"special": true
|
1535 |
+
},
|
1536 |
+
{
|
1537 |
+
"id": 170,
|
1538 |
+
"content": "<SENTINEL_ID_113>",
|
1539 |
+
"single_word": false,
|
1540 |
+
"lstrip": false,
|
1541 |
+
"rstrip": false,
|
1542 |
+
"normalized": false,
|
1543 |
+
"special": true
|
1544 |
+
},
|
1545 |
+
{
|
1546 |
+
"id": 171,
|
1547 |
+
"content": "<SENTINEL_ID_114>",
|
1548 |
+
"single_word": false,
|
1549 |
+
"lstrip": false,
|
1550 |
+
"rstrip": false,
|
1551 |
+
"normalized": false,
|
1552 |
+
"special": true
|
1553 |
+
},
|
1554 |
+
{
|
1555 |
+
"id": 172,
|
1556 |
+
"content": "<SENTINEL_ID_115>",
|
1557 |
+
"single_word": false,
|
1558 |
+
"lstrip": false,
|
1559 |
+
"rstrip": false,
|
1560 |
+
"normalized": false,
|
1561 |
+
"special": true
|
1562 |
+
},
|
1563 |
+
{
|
1564 |
+
"id": 173,
|
1565 |
+
"content": "<SENTINEL_ID_116>",
|
1566 |
+
"single_word": false,
|
1567 |
+
"lstrip": false,
|
1568 |
+
"rstrip": false,
|
1569 |
+
"normalized": false,
|
1570 |
+
"special": true
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"id": 174,
|
1574 |
+
"content": "<SENTINEL_ID_117>",
|
1575 |
+
"single_word": false,
|
1576 |
+
"lstrip": false,
|
1577 |
+
"rstrip": false,
|
1578 |
+
"normalized": false,
|
1579 |
+
"special": true
|
1580 |
+
},
|
1581 |
+
{
|
1582 |
+
"id": 175,
|
1583 |
+
"content": "<SENTINEL_ID_118>",
|
1584 |
+
"single_word": false,
|
1585 |
+
"lstrip": false,
|
1586 |
+
"rstrip": false,
|
1587 |
+
"normalized": false,
|
1588 |
+
"special": true
|
1589 |
+
},
|
1590 |
+
{
|
1591 |
+
"id": 176,
|
1592 |
+
"content": "<SENTINEL_ID_119>",
|
1593 |
+
"single_word": false,
|
1594 |
+
"lstrip": false,
|
1595 |
+
"rstrip": false,
|
1596 |
+
"normalized": false,
|
1597 |
+
"special": true
|
1598 |
+
},
|
1599 |
+
{
|
1600 |
+
"id": 177,
|
1601 |
+
"content": "<SENTINEL_ID_120>",
|
1602 |
+
"single_word": false,
|
1603 |
+
"lstrip": false,
|
1604 |
+
"rstrip": false,
|
1605 |
+
"normalized": false,
|
1606 |
+
"special": true
|
1607 |
+
},
|
1608 |
+
{
|
1609 |
+
"id": 178,
|
1610 |
+
"content": "<SENTINEL_ID_121>",
|
1611 |
+
"single_word": false,
|
1612 |
+
"lstrip": false,
|
1613 |
+
"rstrip": false,
|
1614 |
+
"normalized": false,
|
1615 |
+
"special": true
|
1616 |
+
},
|
1617 |
+
{
|
1618 |
+
"id": 179,
|
1619 |
+
"content": "<SENTINEL_ID_122>",
|
1620 |
+
"single_word": false,
|
1621 |
+
"lstrip": false,
|
1622 |
+
"rstrip": false,
|
1623 |
+
"normalized": false,
|
1624 |
+
"special": true
|
1625 |
+
},
|
1626 |
+
{
|
1627 |
+
"id": 180,
|
1628 |
+
"content": "<SENTINEL_ID_123>",
|
1629 |
+
"single_word": false,
|
1630 |
+
"lstrip": false,
|
1631 |
+
"rstrip": false,
|
1632 |
+
"normalized": false,
|
1633 |
+
"special": true
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"id": 181,
|
1637 |
+
"content": "<SENTINEL_ID_124>",
|
1638 |
+
"single_word": false,
|
1639 |
+
"lstrip": false,
|
1640 |
+
"rstrip": false,
|
1641 |
+
"normalized": false,
|
1642 |
+
"special": true
|
1643 |
+
},
|
1644 |
+
{
|
1645 |
+
"id": 182,
|
1646 |
+
"content": "<SENTINEL_ID_125>",
|
1647 |
+
"single_word": false,
|
1648 |
+
"lstrip": false,
|
1649 |
+
"rstrip": false,
|
1650 |
+
"normalized": false,
|
1651 |
+
"special": true
|
1652 |
+
},
|
1653 |
+
{
|
1654 |
+
"id": 183,
|
1655 |
+
"content": "<SENTINEL_ID_126>",
|
1656 |
+
"single_word": false,
|
1657 |
+
"lstrip": false,
|
1658 |
+
"rstrip": false,
|
1659 |
+
"normalized": false,
|
1660 |
+
"special": true
|
1661 |
+
},
|
1662 |
+
{
|
1663 |
+
"id": 184,
|
1664 |
+
"content": "<SENTINEL_ID_127>",
|
1665 |
+
"single_word": false,
|
1666 |
+
"lstrip": false,
|
1667 |
+
"rstrip": false,
|
1668 |
+
"normalized": false,
|
1669 |
+
"special": true
|
1670 |
+
},
|
1671 |
+
{
|
1672 |
+
"id": 185,
|
1673 |
+
"content": "<SENTINEL_ID_128>",
|
1674 |
+
"single_word": false,
|
1675 |
+
"lstrip": false,
|
1676 |
+
"rstrip": false,
|
1677 |
+
"normalized": false,
|
1678 |
+
"special": true
|
1679 |
+
},
|
1680 |
+
{
|
1681 |
+
"id": 186,
|
1682 |
+
"content": "<SENTINEL_ID_129>",
|
1683 |
+
"single_word": false,
|
1684 |
+
"lstrip": false,
|
1685 |
+
"rstrip": false,
|
1686 |
+
"normalized": false,
|
1687 |
+
"special": true
|
1688 |
+
},
|
1689 |
+
{
|
1690 |
+
"id": 187,
|
1691 |
+
"content": "<SENTINEL_ID_130>",
|
1692 |
+
"single_word": false,
|
1693 |
+
"lstrip": false,
|
1694 |
+
"rstrip": false,
|
1695 |
+
"normalized": false,
|
1696 |
+
"special": true
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"id": 188,
|
1700 |
+
"content": "<SENTINEL_ID_131>",
|
1701 |
+
"single_word": false,
|
1702 |
+
"lstrip": false,
|
1703 |
+
"rstrip": false,
|
1704 |
+
"normalized": false,
|
1705 |
+
"special": true
|
1706 |
+
},
|
1707 |
+
{
|
1708 |
+
"id": 189,
|
1709 |
+
"content": "<SENTINEL_ID_132>",
|
1710 |
+
"single_word": false,
|
1711 |
+
"lstrip": false,
|
1712 |
+
"rstrip": false,
|
1713 |
+
"normalized": false,
|
1714 |
+
"special": true
|
1715 |
+
},
|
1716 |
+
{
|
1717 |
+
"id": 190,
|
1718 |
+
"content": "<SENTINEL_ID_133>",
|
1719 |
+
"single_word": false,
|
1720 |
+
"lstrip": false,
|
1721 |
+
"rstrip": false,
|
1722 |
+
"normalized": false,
|
1723 |
+
"special": true
|
1724 |
+
},
|
1725 |
+
{
|
1726 |
+
"id": 191,
|
1727 |
+
"content": "<SENTINEL_ID_134>",
|
1728 |
+
"single_word": false,
|
1729 |
+
"lstrip": false,
|
1730 |
+
"rstrip": false,
|
1731 |
+
"normalized": false,
|
1732 |
+
"special": true
|
1733 |
+
},
|
1734 |
+
{
|
1735 |
+
"id": 192,
|
1736 |
+
"content": "<SENTINEL_ID_135>",
|
1737 |
+
"single_word": false,
|
1738 |
+
"lstrip": false,
|
1739 |
+
"rstrip": false,
|
1740 |
+
"normalized": false,
|
1741 |
+
"special": true
|
1742 |
+
},
|
1743 |
+
{
|
1744 |
+
"id": 193,
|
1745 |
+
"content": "<SENTINEL_ID_136>",
|
1746 |
+
"single_word": false,
|
1747 |
+
"lstrip": false,
|
1748 |
+
"rstrip": false,
|
1749 |
+
"normalized": false,
|
1750 |
+
"special": true
|
1751 |
+
},
|
1752 |
+
{
|
1753 |
+
"id": 194,
|
1754 |
+
"content": "<SENTINEL_ID_137>",
|
1755 |
+
"single_word": false,
|
1756 |
+
"lstrip": false,
|
1757 |
+
"rstrip": false,
|
1758 |
+
"normalized": false,
|
1759 |
+
"special": true
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"id": 195,
|
1763 |
+
"content": "<SENTINEL_ID_138>",
|
1764 |
+
"single_word": false,
|
1765 |
+
"lstrip": false,
|
1766 |
+
"rstrip": false,
|
1767 |
+
"normalized": false,
|
1768 |
+
"special": true
|
1769 |
+
},
|
1770 |
+
{
|
1771 |
+
"id": 196,
|
1772 |
+
"content": "<SENTINEL_ID_139>",
|
1773 |
+
"single_word": false,
|
1774 |
+
"lstrip": false,
|
1775 |
+
"rstrip": false,
|
1776 |
+
"normalized": false,
|
1777 |
+
"special": true
|
1778 |
+
},
|
1779 |
+
{
|
1780 |
+
"id": 197,
|
1781 |
+
"content": "<SENTINEL_ID_140>",
|
1782 |
+
"single_word": false,
|
1783 |
+
"lstrip": false,
|
1784 |
+
"rstrip": false,
|
1785 |
+
"normalized": false,
|
1786 |
+
"special": true
|
1787 |
+
},
|
1788 |
+
{
|
1789 |
+
"id": 198,
|
1790 |
+
"content": "<SENTINEL_ID_141>",
|
1791 |
+
"single_word": false,
|
1792 |
+
"lstrip": false,
|
1793 |
+
"rstrip": false,
|
1794 |
+
"normalized": false,
|
1795 |
+
"special": true
|
1796 |
+
},
|
1797 |
+
{
|
1798 |
+
"id": 199,
|
1799 |
+
"content": "<SENTINEL_ID_142>",
|
1800 |
+
"single_word": false,
|
1801 |
+
"lstrip": false,
|
1802 |
+
"rstrip": false,
|
1803 |
+
"normalized": false,
|
1804 |
+
"special": true
|
1805 |
+
},
|
1806 |
+
{
|
1807 |
+
"id": 200,
|
1808 |
+
"content": "<SENTINEL_ID_143>",
|
1809 |
+
"single_word": false,
|
1810 |
+
"lstrip": false,
|
1811 |
+
"rstrip": false,
|
1812 |
+
"normalized": false,
|
1813 |
+
"special": true
|
1814 |
+
},
|
1815 |
+
{
|
1816 |
+
"id": 201,
|
1817 |
+
"content": "<SENTINEL_ID_144>",
|
1818 |
+
"single_word": false,
|
1819 |
+
"lstrip": false,
|
1820 |
+
"rstrip": false,
|
1821 |
+
"normalized": false,
|
1822 |
+
"special": true
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"id": 202,
|
1826 |
+
"content": "<SENTINEL_ID_145>",
|
1827 |
+
"single_word": false,
|
1828 |
+
"lstrip": false,
|
1829 |
+
"rstrip": false,
|
1830 |
+
"normalized": false,
|
1831 |
+
"special": true
|
1832 |
+
},
|
1833 |
+
{
|
1834 |
+
"id": 203,
|
1835 |
+
"content": "<SENTINEL_ID_146>",
|
1836 |
+
"single_word": false,
|
1837 |
+
"lstrip": false,
|
1838 |
+
"rstrip": false,
|
1839 |
+
"normalized": false,
|
1840 |
+
"special": true
|
1841 |
+
},
|
1842 |
+
{
|
1843 |
+
"id": 204,
|
1844 |
+
"content": "<SENTINEL_ID_147>",
|
1845 |
+
"single_word": false,
|
1846 |
+
"lstrip": false,
|
1847 |
+
"rstrip": false,
|
1848 |
+
"normalized": false,
|
1849 |
+
"special": true
|
1850 |
+
},
|
1851 |
+
{
|
1852 |
+
"id": 205,
|
1853 |
+
"content": "<SENTINEL_ID_148>",
|
1854 |
+
"single_word": false,
|
1855 |
+
"lstrip": false,
|
1856 |
+
"rstrip": false,
|
1857 |
+
"normalized": false,
|
1858 |
+
"special": true
|
1859 |
+
},
|
1860 |
+
{
|
1861 |
+
"id": 206,
|
1862 |
+
"content": "<SENTINEL_ID_149>",
|
1863 |
+
"single_word": false,
|
1864 |
+
"lstrip": false,
|
1865 |
+
"rstrip": false,
|
1866 |
+
"normalized": false,
|
1867 |
+
"special": true
|
1868 |
+
},
|
1869 |
+
{
|
1870 |
+
"id": 207,
|
1871 |
+
"content": "<SENTINEL_ID_150>",
|
1872 |
+
"single_word": false,
|
1873 |
+
"lstrip": false,
|
1874 |
+
"rstrip": false,
|
1875 |
+
"normalized": false,
|
1876 |
+
"special": true
|
1877 |
+
},
|
1878 |
+
{
|
1879 |
+
"id": 208,
|
1880 |
+
"content": "<SENTINEL_ID_151>",
|
1881 |
+
"single_word": false,
|
1882 |
+
"lstrip": false,
|
1883 |
+
"rstrip": false,
|
1884 |
+
"normalized": false,
|
1885 |
+
"special": true
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"id": 209,
|
1889 |
+
"content": "<SENTINEL_ID_152>",
|
1890 |
+
"single_word": false,
|
1891 |
+
"lstrip": false,
|
1892 |
+
"rstrip": false,
|
1893 |
+
"normalized": false,
|
1894 |
+
"special": true
|
1895 |
+
},
|
1896 |
+
{
|
1897 |
+
"id": 210,
|
1898 |
+
"content": "<SENTINEL_ID_153>",
|
1899 |
+
"single_word": false,
|
1900 |
+
"lstrip": false,
|
1901 |
+
"rstrip": false,
|
1902 |
+
"normalized": false,
|
1903 |
+
"special": true
|
1904 |
+
},
|
1905 |
+
{
|
1906 |
+
"id": 211,
|
1907 |
+
"content": "<SENTINEL_ID_154>",
|
1908 |
+
"single_word": false,
|
1909 |
+
"lstrip": false,
|
1910 |
+
"rstrip": false,
|
1911 |
+
"normalized": false,
|
1912 |
+
"special": true
|
1913 |
+
},
|
1914 |
+
{
|
1915 |
+
"id": 212,
|
1916 |
+
"content": "<SENTINEL_ID_155>",
|
1917 |
+
"single_word": false,
|
1918 |
+
"lstrip": false,
|
1919 |
+
"rstrip": false,
|
1920 |
+
"normalized": false,
|
1921 |
+
"special": true
|
1922 |
+
},
|
1923 |
+
{
|
1924 |
+
"id": 213,
|
1925 |
+
"content": "<SENTINEL_ID_156>",
|
1926 |
+
"single_word": false,
|
1927 |
+
"lstrip": false,
|
1928 |
+
"rstrip": false,
|
1929 |
+
"normalized": false,
|
1930 |
+
"special": true
|
1931 |
+
},
|
1932 |
+
{
|
1933 |
+
"id": 214,
|
1934 |
+
"content": "<SENTINEL_ID_157>",
|
1935 |
+
"single_word": false,
|
1936 |
+
"lstrip": false,
|
1937 |
+
"rstrip": false,
|
1938 |
+
"normalized": false,
|
1939 |
+
"special": true
|
1940 |
+
},
|
1941 |
+
{
|
1942 |
+
"id": 215,
|
1943 |
+
"content": "<SENTINEL_ID_158>",
|
1944 |
+
"single_word": false,
|
1945 |
+
"lstrip": false,
|
1946 |
+
"rstrip": false,
|
1947 |
+
"normalized": false,
|
1948 |
+
"special": true
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"id": 216,
|
1952 |
+
"content": "<SENTINEL_ID_159>",
|
1953 |
+
"single_word": false,
|
1954 |
+
"lstrip": false,
|
1955 |
+
"rstrip": false,
|
1956 |
+
"normalized": false,
|
1957 |
+
"special": true
|
1958 |
+
},
|
1959 |
+
{
|
1960 |
+
"id": 217,
|
1961 |
+
"content": "<SENTINEL_ID_160>",
|
1962 |
+
"single_word": false,
|
1963 |
+
"lstrip": false,
|
1964 |
+
"rstrip": false,
|
1965 |
+
"normalized": false,
|
1966 |
+
"special": true
|
1967 |
+
},
|
1968 |
+
{
|
1969 |
+
"id": 218,
|
1970 |
+
"content": "<SENTINEL_ID_161>",
|
1971 |
+
"single_word": false,
|
1972 |
+
"lstrip": false,
|
1973 |
+
"rstrip": false,
|
1974 |
+
"normalized": false,
|
1975 |
+
"special": true
|
1976 |
+
},
|
1977 |
+
{
|
1978 |
+
"id": 219,
|
1979 |
+
"content": "<SENTINEL_ID_162>",
|
1980 |
+
"single_word": false,
|
1981 |
+
"lstrip": false,
|
1982 |
+
"rstrip": false,
|
1983 |
+
"normalized": false,
|
1984 |
+
"special": true
|
1985 |
+
},
|
1986 |
+
{
|
1987 |
+
"id": 220,
|
1988 |
+
"content": "<SENTINEL_ID_163>",
|
1989 |
+
"single_word": false,
|
1990 |
+
"lstrip": false,
|
1991 |
+
"rstrip": false,
|
1992 |
+
"normalized": false,
|
1993 |
+
"special": true
|
1994 |
+
},
|
1995 |
+
{
|
1996 |
+
"id": 221,
|
1997 |
+
"content": "<SENTINEL_ID_164>",
|
1998 |
+
"single_word": false,
|
1999 |
+
"lstrip": false,
|
2000 |
+
"rstrip": false,
|
2001 |
+
"normalized": false,
|
2002 |
+
"special": true
|
2003 |
+
},
|
2004 |
+
{
|
2005 |
+
"id": 222,
|
2006 |
+
"content": "<SENTINEL_ID_165>",
|
2007 |
+
"single_word": false,
|
2008 |
+
"lstrip": false,
|
2009 |
+
"rstrip": false,
|
2010 |
+
"normalized": false,
|
2011 |
+
"special": true
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"id": 223,
|
2015 |
+
"content": "<SENTINEL_ID_166>",
|
2016 |
+
"single_word": false,
|
2017 |
+
"lstrip": false,
|
2018 |
+
"rstrip": false,
|
2019 |
+
"normalized": false,
|
2020 |
+
"special": true
|
2021 |
+
},
|
2022 |
+
{
|
2023 |
+
"id": 224,
|
2024 |
+
"content": "<SENTINEL_ID_167>",
|
2025 |
+
"single_word": false,
|
2026 |
+
"lstrip": false,
|
2027 |
+
"rstrip": false,
|
2028 |
+
"normalized": false,
|
2029 |
+
"special": true
|
2030 |
+
},
|
2031 |
+
{
|
2032 |
+
"id": 225,
|
2033 |
+
"content": "<SENTINEL_ID_168>",
|
2034 |
+
"single_word": false,
|
2035 |
+
"lstrip": false,
|
2036 |
+
"rstrip": false,
|
2037 |
+
"normalized": false,
|
2038 |
+
"special": true
|
2039 |
+
},
|
2040 |
+
{
|
2041 |
+
"id": 226,
|
2042 |
+
"content": "<SENTINEL_ID_169>",
|
2043 |
+
"single_word": false,
|
2044 |
+
"lstrip": false,
|
2045 |
+
"rstrip": false,
|
2046 |
+
"normalized": false,
|
2047 |
+
"special": true
|
2048 |
+
},
|
2049 |
+
{
|
2050 |
+
"id": 227,
|
2051 |
+
"content": "<SENTINEL_ID_170>",
|
2052 |
+
"single_word": false,
|
2053 |
+
"lstrip": false,
|
2054 |
+
"rstrip": false,
|
2055 |
+
"normalized": false,
|
2056 |
+
"special": true
|
2057 |
+
},
|
2058 |
+
{
|
2059 |
+
"id": 228,
|
2060 |
+
"content": "<SENTINEL_ID_171>",
|
2061 |
+
"single_word": false,
|
2062 |
+
"lstrip": false,
|
2063 |
+
"rstrip": false,
|
2064 |
+
"normalized": false,
|
2065 |
+
"special": true
|
2066 |
+
},
|
2067 |
+
{
|
2068 |
+
"id": 229,
|
2069 |
+
"content": "<SENTINEL_ID_172>",
|
2070 |
+
"single_word": false,
|
2071 |
+
"lstrip": false,
|
2072 |
+
"rstrip": false,
|
2073 |
+
"normalized": false,
|
2074 |
+
"special": true
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"id": 230,
|
2078 |
+
"content": "<SENTINEL_ID_173>",
|
2079 |
+
"single_word": false,
|
2080 |
+
"lstrip": false,
|
2081 |
+
"rstrip": false,
|
2082 |
+
"normalized": false,
|
2083 |
+
"special": true
|
2084 |
+
},
|
2085 |
+
{
|
2086 |
+
"id": 231,
|
2087 |
+
"content": "<SENTINEL_ID_174>",
|
2088 |
+
"single_word": false,
|
2089 |
+
"lstrip": false,
|
2090 |
+
"rstrip": false,
|
2091 |
+
"normalized": false,
|
2092 |
+
"special": true
|
2093 |
+
},
|
2094 |
+
{
|
2095 |
+
"id": 232,
|
2096 |
+
"content": "<SENTINEL_ID_175>",
|
2097 |
+
"single_word": false,
|
2098 |
+
"lstrip": false,
|
2099 |
+
"rstrip": false,
|
2100 |
+
"normalized": false,
|
2101 |
+
"special": true
|
2102 |
+
},
|
2103 |
+
{
|
2104 |
+
"id": 233,
|
2105 |
+
"content": "<SENTINEL_ID_176>",
|
2106 |
+
"single_word": false,
|
2107 |
+
"lstrip": false,
|
2108 |
+
"rstrip": false,
|
2109 |
+
"normalized": false,
|
2110 |
+
"special": true
|
2111 |
+
},
|
2112 |
+
{
|
2113 |
+
"id": 234,
|
2114 |
+
"content": "<SENTINEL_ID_177>",
|
2115 |
+
"single_word": false,
|
2116 |
+
"lstrip": false,
|
2117 |
+
"rstrip": false,
|
2118 |
+
"normalized": false,
|
2119 |
+
"special": true
|
2120 |
+
},
|
2121 |
+
{
|
2122 |
+
"id": 235,
|
2123 |
+
"content": "<SENTINEL_ID_178>",
|
2124 |
+
"single_word": false,
|
2125 |
+
"lstrip": false,
|
2126 |
+
"rstrip": false,
|
2127 |
+
"normalized": false,
|
2128 |
+
"special": true
|
2129 |
+
},
|
2130 |
+
{
|
2131 |
+
"id": 236,
|
2132 |
+
"content": "<SENTINEL_ID_179>",
|
2133 |
+
"single_word": false,
|
2134 |
+
"lstrip": false,
|
2135 |
+
"rstrip": false,
|
2136 |
+
"normalized": false,
|
2137 |
+
"special": true
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"id": 237,
|
2141 |
+
"content": "<SENTINEL_ID_180>",
|
2142 |
+
"single_word": false,
|
2143 |
+
"lstrip": false,
|
2144 |
+
"rstrip": false,
|
2145 |
+
"normalized": false,
|
2146 |
+
"special": true
|
2147 |
+
},
|
2148 |
+
{
|
2149 |
+
"id": 238,
|
2150 |
+
"content": "<SENTINEL_ID_181>",
|
2151 |
+
"single_word": false,
|
2152 |
+
"lstrip": false,
|
2153 |
+
"rstrip": false,
|
2154 |
+
"normalized": false,
|
2155 |
+
"special": true
|
2156 |
+
},
|
2157 |
+
{
|
2158 |
+
"id": 239,
|
2159 |
+
"content": "<SENTINEL_ID_182>",
|
2160 |
+
"single_word": false,
|
2161 |
+
"lstrip": false,
|
2162 |
+
"rstrip": false,
|
2163 |
+
"normalized": false,
|
2164 |
+
"special": true
|
2165 |
+
},
|
2166 |
+
{
|
2167 |
+
"id": 240,
|
2168 |
+
"content": "<SENTINEL_ID_183>",
|
2169 |
+
"single_word": false,
|
2170 |
+
"lstrip": false,
|
2171 |
+
"rstrip": false,
|
2172 |
+
"normalized": false,
|
2173 |
+
"special": true
|
2174 |
+
},
|
2175 |
+
{
|
2176 |
+
"id": 241,
|
2177 |
+
"content": "<SENTINEL_ID_184>",
|
2178 |
+
"single_word": false,
|
2179 |
+
"lstrip": false,
|
2180 |
+
"rstrip": false,
|
2181 |
+
"normalized": false,
|
2182 |
+
"special": true
|
2183 |
+
},
|
2184 |
+
{
|
2185 |
+
"id": 242,
|
2186 |
+
"content": "<SENTINEL_ID_185>",
|
2187 |
+
"single_word": false,
|
2188 |
+
"lstrip": false,
|
2189 |
+
"rstrip": false,
|
2190 |
+
"normalized": false,
|
2191 |
+
"special": true
|
2192 |
+
},
|
2193 |
+
{
|
2194 |
+
"id": 243,
|
2195 |
+
"content": "<SENTINEL_ID_186>",
|
2196 |
+
"single_word": false,
|
2197 |
+
"lstrip": false,
|
2198 |
+
"rstrip": false,
|
2199 |
+
"normalized": false,
|
2200 |
+
"special": true
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"id": 244,
|
2204 |
+
"content": "<SENTINEL_ID_187>",
|
2205 |
+
"single_word": false,
|
2206 |
+
"lstrip": false,
|
2207 |
+
"rstrip": false,
|
2208 |
+
"normalized": false,
|
2209 |
+
"special": true
|
2210 |
+
},
|
2211 |
+
{
|
2212 |
+
"id": 245,
|
2213 |
+
"content": "<SENTINEL_ID_188>",
|
2214 |
+
"single_word": false,
|
2215 |
+
"lstrip": false,
|
2216 |
+
"rstrip": false,
|
2217 |
+
"normalized": false,
|
2218 |
+
"special": true
|
2219 |
+
},
|
2220 |
+
{
|
2221 |
+
"id": 246,
|
2222 |
+
"content": "<SENTINEL_ID_189>",
|
2223 |
+
"single_word": false,
|
2224 |
+
"lstrip": false,
|
2225 |
+
"rstrip": false,
|
2226 |
+
"normalized": false,
|
2227 |
+
"special": true
|
2228 |
+
},
|
2229 |
+
{
|
2230 |
+
"id": 247,
|
2231 |
+
"content": "<SENTINEL_ID_190>",
|
2232 |
+
"single_word": false,
|
2233 |
+
"lstrip": false,
|
2234 |
+
"rstrip": false,
|
2235 |
+
"normalized": false,
|
2236 |
+
"special": true
|
2237 |
+
},
|
2238 |
+
{
|
2239 |
+
"id": 248,
|
2240 |
+
"content": "<SENTINEL_ID_191>",
|
2241 |
+
"single_word": false,
|
2242 |
+
"lstrip": false,
|
2243 |
+
"rstrip": false,
|
2244 |
+
"normalized": false,
|
2245 |
+
"special": true
|
2246 |
+
},
|
2247 |
+
{
|
2248 |
+
"id": 249,
|
2249 |
+
"content": "<SENTINEL_ID_192>",
|
2250 |
+
"single_word": false,
|
2251 |
+
"lstrip": false,
|
2252 |
+
"rstrip": false,
|
2253 |
+
"normalized": false,
|
2254 |
+
"special": true
|
2255 |
+
},
|
2256 |
+
{
|
2257 |
+
"id": 250,
|
2258 |
+
"content": "<SENTINEL_ID_193>",
|
2259 |
+
"single_word": false,
|
2260 |
+
"lstrip": false,
|
2261 |
+
"rstrip": false,
|
2262 |
+
"normalized": false,
|
2263 |
+
"special": true
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"id": 251,
|
2267 |
+
"content": "<SENTINEL_ID_194>",
|
2268 |
+
"single_word": false,
|
2269 |
+
"lstrip": false,
|
2270 |
+
"rstrip": false,
|
2271 |
+
"normalized": false,
|
2272 |
+
"special": true
|
2273 |
+
},
|
2274 |
+
{
|
2275 |
+
"id": 252,
|
2276 |
+
"content": "<SENTINEL_ID_195>",
|
2277 |
+
"single_word": false,
|
2278 |
+
"lstrip": false,
|
2279 |
+
"rstrip": false,
|
2280 |
+
"normalized": false,
|
2281 |
+
"special": true
|
2282 |
+
},
|
2283 |
+
{
|
2284 |
+
"id": 253,
|
2285 |
+
"content": "<SENTINEL_ID_196>",
|
2286 |
+
"single_word": false,
|
2287 |
+
"lstrip": false,
|
2288 |
+
"rstrip": false,
|
2289 |
+
"normalized": false,
|
2290 |
+
"special": true
|
2291 |
+
},
|
2292 |
+
{
|
2293 |
+
"id": 254,
|
2294 |
+
"content": "<SENTINEL_ID_197>",
|
2295 |
+
"single_word": false,
|
2296 |
+
"lstrip": false,
|
2297 |
+
"rstrip": false,
|
2298 |
+
"normalized": false,
|
2299 |
+
"special": true
|
2300 |
+
},
|
2301 |
+
{
|
2302 |
+
"id": 255,
|
2303 |
+
"content": "<SENTINEL_ID_198>",
|
2304 |
+
"single_word": false,
|
2305 |
+
"lstrip": false,
|
2306 |
+
"rstrip": false,
|
2307 |
+
"normalized": false,
|
2308 |
+
"special": true
|
2309 |
+
},
|
2310 |
+
{
|
2311 |
+
"id": 256,
|
2312 |
+
"content": "<SENTINEL_ID_199>",
|
2313 |
+
"single_word": false,
|
2314 |
+
"lstrip": false,
|
2315 |
+
"rstrip": false,
|
2316 |
+
"normalized": false,
|
2317 |
+
"special": true
|
2318 |
+
},
|
2319 |
+
{
|
2320 |
+
"id": 257,
|
2321 |
+
"content": "<MOLECULAR_ENTITY_TYPE_ANTIGEN>",
|
2322 |
+
"single_word": false,
|
2323 |
+
"lstrip": false,
|
2324 |
+
"rstrip": false,
|
2325 |
+
"normalized": false,
|
2326 |
+
"special": true
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"id": 258,
|
2330 |
+
"content": "<MOLECULAR_ENTITY_TYPE_ANTIBODY_LIGHT_CHAIN>",
|
2331 |
+
"single_word": false,
|
2332 |
+
"lstrip": false,
|
2333 |
+
"rstrip": false,
|
2334 |
+
"normalized": false,
|
2335 |
+
"special": true
|
2336 |
+
},
|
2337 |
+
{
|
2338 |
+
"id": 259,
|
2339 |
+
"content": "<MOLECULAR_ENTITY_TYPE_ANTIBODY_HEAVY_CHAIN>",
|
2340 |
+
"single_word": false,
|
2341 |
+
"lstrip": false,
|
2342 |
+
"rstrip": false,
|
2343 |
+
"normalized": false,
|
2344 |
+
"special": true
|
2345 |
+
},
|
2346 |
+
{
|
2347 |
+
"id": 260,
|
2348 |
+
"content": "<ATTRIBUTE_ORGANISM>",
|
2349 |
+
"single_word": false,
|
2350 |
+
"lstrip": false,
|
2351 |
+
"rstrip": false,
|
2352 |
+
"normalized": false,
|
2353 |
+
"special": true
|
2354 |
+
},
|
2355 |
+
{
|
2356 |
+
"id": 261,
|
2357 |
+
"content": "<ATTRIBUTE_ORGANISM_HUMAN>",
|
2358 |
+
"single_word": false,
|
2359 |
+
"lstrip": false,
|
2360 |
+
"rstrip": false,
|
2361 |
+
"normalized": false,
|
2362 |
+
"special": true
|
2363 |
+
},
|
2364 |
+
{
|
2365 |
+
"id": 262,
|
2366 |
+
"content": "<ATTRIBUTE_ORGANISM_RABBIT>",
|
2367 |
+
"single_word": false,
|
2368 |
+
"lstrip": false,
|
2369 |
+
"rstrip": false,
|
2370 |
+
"normalized": false,
|
2371 |
+
"special": true
|
2372 |
+
},
|
2373 |
+
{
|
2374 |
+
"id": 263,
|
2375 |
+
"content": "<ATTRIBUTE_ORGANISM_RAT>",
|
2376 |
+
"single_word": false,
|
2377 |
+
"lstrip": false,
|
2378 |
+
"rstrip": false,
|
2379 |
+
"normalized": false,
|
2380 |
+
"special": true
|
2381 |
+
},
|
2382 |
+
{
|
2383 |
+
"id": 264,
|
2384 |
+
"content": "<ATTRIBUTE_ORGANISM_MOUSE>",
|
2385 |
+
"single_word": false,
|
2386 |
+
"lstrip": false,
|
2387 |
+
"rstrip": false,
|
2388 |
+
"normalized": false,
|
2389 |
+
"special": true
|
2390 |
+
},
|
2391 |
+
{
|
2392 |
+
"id": 265,
|
2393 |
+
"content": "<ATTRIBUTE_ORGANISM_MONKEY>",
|
2394 |
+
"single_word": false,
|
2395 |
+
"lstrip": false,
|
2396 |
+
"rstrip": false,
|
2397 |
+
"normalized": false,
|
2398 |
+
"special": true
|
2399 |
+
},
|
2400 |
+
{
|
2401 |
+
"id": 266,
|
2402 |
+
"content": "<ATTRIBUTE_ORGANISM_CAMEL>",
|
2403 |
+
"single_word": false,
|
2404 |
+
"lstrip": false,
|
2405 |
+
"rstrip": false,
|
2406 |
+
"normalized": false,
|
2407 |
+
"special": true
|
2408 |
+
},
|
2409 |
+
{
|
2410 |
+
"id": 267,
|
2411 |
+
"content": "<EPITOPE_PARATOPE_PREDICTION>",
|
2412 |
+
"single_word": false,
|
2413 |
+
"lstrip": false,
|
2414 |
+
"rstrip": false,
|
2415 |
+
"normalized": false,
|
2416 |
+
"special": true
|
2417 |
+
},
|
2418 |
+
{
|
2419 |
+
"id": 268,
|
2420 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN_CDR1>",
|
2421 |
+
"single_word": false,
|
2422 |
+
"lstrip": false,
|
2423 |
+
"rstrip": false,
|
2424 |
+
"normalized": false,
|
2425 |
+
"special": true
|
2426 |
+
},
|
2427 |
+
{
|
2428 |
+
"id": 269,
|
2429 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN_CDR3>",
|
2430 |
+
"single_word": false,
|
2431 |
+
"lstrip": false,
|
2432 |
+
"rstrip": false,
|
2433 |
+
"normalized": false,
|
2434 |
+
"special": true
|
2435 |
+
},
|
2436 |
+
{
|
2437 |
+
"id": 270,
|
2438 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN_CDR3>",
|
2439 |
+
"single_word": false,
|
2440 |
+
"lstrip": false,
|
2441 |
+
"rstrip": false,
|
2442 |
+
"normalized": false,
|
2443 |
+
"special": true
|
2444 |
+
},
|
2445 |
+
{
|
2446 |
+
"id": 271,
|
2447 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN_CDR2>",
|
2448 |
+
"single_word": false,
|
2449 |
+
"lstrip": false,
|
2450 |
+
"rstrip": false,
|
2451 |
+
"normalized": false,
|
2452 |
+
"special": true
|
2453 |
+
},
|
2454 |
+
{
|
2455 |
+
"id": 272,
|
2456 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN_CDR2>",
|
2457 |
+
"single_word": false,
|
2458 |
+
"lstrip": false,
|
2459 |
+
"rstrip": false,
|
2460 |
+
"normalized": false,
|
2461 |
+
"special": true
|
2462 |
+
},
|
2463 |
+
{
|
2464 |
+
"id": 273,
|
2465 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN_CDR1>",
|
2466 |
+
"single_word": false,
|
2467 |
+
"lstrip": false,
|
2468 |
+
"rstrip": false,
|
2469 |
+
"normalized": false,
|
2470 |
+
"special": true
|
2471 |
+
},
|
2472 |
+
{
|
2473 |
+
"id": 274,
|
2474 |
+
"content": "<MOLECULAR_ENTITY_GENERAL_PROTEIN>",
|
2475 |
+
"single_word": false,
|
2476 |
+
"lstrip": false,
|
2477 |
+
"rstrip": false,
|
2478 |
+
"normalized": false,
|
2479 |
+
"special": true
|
2480 |
+
},
|
2481 |
+
{
|
2482 |
+
"id": 275,
|
2483 |
+
"content": "<TIMESTEP>",
|
2484 |
+
"single_word": false,
|
2485 |
+
"lstrip": false,
|
2486 |
+
"rstrip": false,
|
2487 |
+
"normalized": false,
|
2488 |
+
"special": true
|
2489 |
+
},
|
2490 |
+
{
|
2491 |
+
"id": 276,
|
2492 |
+
"content": "<DIFFUSION>",
|
2493 |
+
"single_word": false,
|
2494 |
+
"lstrip": false,
|
2495 |
+
"rstrip": false,
|
2496 |
+
"normalized": false,
|
2497 |
+
"special": true
|
2498 |
+
},
|
2499 |
+
{
|
2500 |
+
"id": 277,
|
2501 |
+
"content": "<SEQUENCE_NATURAL_END>",
|
2502 |
+
"single_word": false,
|
2503 |
+
"lstrip": false,
|
2504 |
+
"rstrip": false,
|
2505 |
+
"normalized": false,
|
2506 |
+
"special": true
|
2507 |
+
},
|
2508 |
+
{
|
2509 |
+
"id": 278,
|
2510 |
+
"content": "<SMILES_SEQUENCE>",
|
2511 |
+
"single_word": false,
|
2512 |
+
"lstrip": false,
|
2513 |
+
"rstrip": false,
|
2514 |
+
"normalized": false,
|
2515 |
+
"special": true
|
2516 |
+
},
|
2517 |
+
{
|
2518 |
+
"id": 279,
|
2519 |
+
"content": "<SELFIES_SEQUENCE>",
|
2520 |
+
"single_word": false,
|
2521 |
+
"lstrip": false,
|
2522 |
+
"rstrip": false,
|
2523 |
+
"normalized": false,
|
2524 |
+
"special": true
|
2525 |
+
},
|
2526 |
+
{
|
2527 |
+
"id": 280,
|
2528 |
+
"content": "<AMINO_ACID_SEQUENCE>",
|
2529 |
+
"single_word": false,
|
2530 |
+
"lstrip": false,
|
2531 |
+
"rstrip": false,
|
2532 |
+
"normalized": false,
|
2533 |
+
"special": true
|
2534 |
+
},
|
2535 |
+
{
|
2536 |
+
"id": 281,
|
2537 |
+
"content": "<GENERAL_AFFINITY_CLASS>",
|
2538 |
+
"single_word": false,
|
2539 |
+
"lstrip": false,
|
2540 |
+
"rstrip": false,
|
2541 |
+
"normalized": false,
|
2542 |
+
"special": true
|
2543 |
+
},
|
2544 |
+
{
|
2545 |
+
"id": 282,
|
2546 |
+
"content": "<BACKSPACE>",
|
2547 |
+
"single_word": false,
|
2548 |
+
"lstrip": false,
|
2549 |
+
"rstrip": false,
|
2550 |
+
"normalized": false,
|
2551 |
+
"special": true
|
2552 |
+
},
|
2553 |
+
{
|
2554 |
+
"id": 283,
|
2555 |
+
"content": "<SEQUENCE_NATURAL_START>",
|
2556 |
+
"single_word": false,
|
2557 |
+
"lstrip": false,
|
2558 |
+
"rstrip": false,
|
2559 |
+
"normalized": false,
|
2560 |
+
"special": true
|
2561 |
+
},
|
2562 |
+
{
|
2563 |
+
"id": 284,
|
2564 |
+
"content": "<NOOP>",
|
2565 |
+
"single_word": false,
|
2566 |
+
"lstrip": false,
|
2567 |
+
"rstrip": false,
|
2568 |
+
"normalized": false,
|
2569 |
+
"special": true
|
2570 |
+
},
|
2571 |
+
{
|
2572 |
+
"id": 285,
|
2573 |
+
"content": "<TARGETED_ANTIBODY_DESIGN_ENCODER_ONLY_MODE>",
|
2574 |
+
"single_word": false,
|
2575 |
+
"lstrip": false,
|
2576 |
+
"rstrip": false,
|
2577 |
+
"normalized": false,
|
2578 |
+
"special": true
|
2579 |
+
},
|
2580 |
+
{
|
2581 |
+
"id": 286,
|
2582 |
+
"content": "<MOLECULAR_ENTITY_SMALL_MOLECULE>",
|
2583 |
+
"single_word": false,
|
2584 |
+
"lstrip": false,
|
2585 |
+
"rstrip": false,
|
2586 |
+
"normalized": false,
|
2587 |
+
"special": true
|
2588 |
+
},
|
2589 |
+
{
|
2590 |
+
"id": 287,
|
2591 |
+
"content": "<MOLECULAR_ENTITY_CELL_GENE_EXPRESSION_RANKED>",
|
2592 |
+
"single_word": false,
|
2593 |
+
"lstrip": false,
|
2594 |
+
"rstrip": false,
|
2595 |
+
"normalized": false,
|
2596 |
+
"special": true
|
2597 |
+
},
|
2598 |
+
{
|
2599 |
+
"id": 288,
|
2600 |
+
"content": "<CELL_TYPE_CLASS>",
|
2601 |
+
"single_word": false,
|
2602 |
+
"lstrip": false,
|
2603 |
+
"rstrip": false,
|
2604 |
+
"normalized": false,
|
2605 |
+
"special": true
|
2606 |
+
},
|
2607 |
+
{
|
2608 |
+
"id": 289,
|
2609 |
+
"content": "<TISSUE_TYPE_CLASS>",
|
2610 |
+
"single_word": false,
|
2611 |
+
"lstrip": false,
|
2612 |
+
"rstrip": false,
|
2613 |
+
"normalized": false,
|
2614 |
+
"special": true
|
2615 |
+
},
|
2616 |
+
{
|
2617 |
+
"id": 290,
|
2618 |
+
"content": "<CORRUPTED_AREA_START>",
|
2619 |
+
"single_word": false,
|
2620 |
+
"lstrip": false,
|
2621 |
+
"rstrip": false,
|
2622 |
+
"normalized": false,
|
2623 |
+
"special": true
|
2624 |
+
},
|
2625 |
+
{
|
2626 |
+
"id": 291,
|
2627 |
+
"content": "<CORRUPTED_AREA_END>",
|
2628 |
+
"single_word": false,
|
2629 |
+
"lstrip": false,
|
2630 |
+
"rstrip": false,
|
2631 |
+
"normalized": false,
|
2632 |
+
"special": true
|
2633 |
+
},
|
2634 |
+
{
|
2635 |
+
"id": 292,
|
2636 |
+
"content": "<MOLECULAR_ENTITY_MUTATED_PROTEIN_CHAIN>",
|
2637 |
+
"single_word": false,
|
2638 |
+
"lstrip": false,
|
2639 |
+
"rstrip": false,
|
2640 |
+
"normalized": false,
|
2641 |
+
"special": true
|
2642 |
+
},
|
2643 |
+
{
|
2644 |
+
"id": 293,
|
2645 |
+
"content": "<MOLECULAR_ENTITY_PROTEIN_CHAIN>",
|
2646 |
+
"single_word": false,
|
2647 |
+
"lstrip": false,
|
2648 |
+
"rstrip": false,
|
2649 |
+
"normalized": false,
|
2650 |
+
"special": true
|
2651 |
+
},
|
2652 |
+
{
|
2653 |
+
"id": 294,
|
2654 |
+
"content": "<COMPLEX_ENTITY>",
|
2655 |
+
"single_word": false,
|
2656 |
+
"lstrip": false,
|
2657 |
+
"rstrip": false,
|
2658 |
+
"normalized": false,
|
2659 |
+
"special": true
|
2660 |
+
},
|
2661 |
+
{
|
2662 |
+
"id": 295,
|
2663 |
+
"content": "<ALTERNATIVE>",
|
2664 |
+
"single_word": false,
|
2665 |
+
"lstrip": false,
|
2666 |
+
"rstrip": false,
|
2667 |
+
"normalized": false,
|
2668 |
+
"special": true
|
2669 |
+
},
|
2670 |
+
{
|
2671 |
+
"id": 296,
|
2672 |
+
"content": "<CDR3_REGION>",
|
2673 |
+
"single_word": false,
|
2674 |
+
"lstrip": false,
|
2675 |
+
"rstrip": false,
|
2676 |
+
"normalized": false,
|
2677 |
+
"special": true
|
2678 |
+
},
|
2679 |
+
{
|
2680 |
+
"id": 297,
|
2681 |
+
"content": "<GENERAL_CHAIN>",
|
2682 |
+
"single_word": false,
|
2683 |
+
"lstrip": false,
|
2684 |
+
"rstrip": false,
|
2685 |
+
"normalized": false,
|
2686 |
+
"special": true
|
2687 |
+
},
|
2688 |
+
{
|
2689 |
+
"id": 298,
|
2690 |
+
"content": "<SUBMOLECULAR_ENTITY>",
|
2691 |
+
"single_word": false,
|
2692 |
+
"lstrip": false,
|
2693 |
+
"rstrip": false,
|
2694 |
+
"normalized": false,
|
2695 |
+
"special": true
|
2696 |
+
},
|
2697 |
+
{
|
2698 |
+
"id": 299,
|
2699 |
+
"content": "<MUTATED>",
|
2700 |
+
"single_word": false,
|
2701 |
+
"lstrip": false,
|
2702 |
+
"rstrip": false,
|
2703 |
+
"normalized": false,
|
2704 |
+
"special": true
|
2705 |
+
},
|
2706 |
+
{
|
2707 |
+
"id": 300,
|
2708 |
+
"content": "<MOLECULAR_ENTITY_TCR_ALPHA_CDR3>",
|
2709 |
+
"single_word": false,
|
2710 |
+
"lstrip": false,
|
2711 |
+
"rstrip": false,
|
2712 |
+
"normalized": false,
|
2713 |
+
"special": true
|
2714 |
+
},
|
2715 |
+
{
|
2716 |
+
"id": 301,
|
2717 |
+
"content": "<MOLECULAR_ENTITY_TCR_DELTA_CDR3>",
|
2718 |
+
"single_word": false,
|
2719 |
+
"lstrip": false,
|
2720 |
+
"rstrip": false,
|
2721 |
+
"normalized": false,
|
2722 |
+
"special": true
|
2723 |
+
},
|
2724 |
+
{
|
2725 |
+
"id": 302,
|
2726 |
+
"content": "<MOLECULAR_ENTITY_TCR_DELTA_VAR>",
|
2727 |
+
"single_word": false,
|
2728 |
+
"lstrip": false,
|
2729 |
+
"rstrip": false,
|
2730 |
+
"normalized": false,
|
2731 |
+
"special": true
|
2732 |
+
},
|
2733 |
+
{
|
2734 |
+
"id": 303,
|
2735 |
+
"content": "<MOLECULAR_ENTITY_TCR_GAMMA_CDR3>",
|
2736 |
+
"single_word": false,
|
2737 |
+
"lstrip": false,
|
2738 |
+
"rstrip": false,
|
2739 |
+
"normalized": false,
|
2740 |
+
"special": true
|
2741 |
+
},
|
2742 |
+
{
|
2743 |
+
"id": 304,
|
2744 |
+
"content": "<MOLECULAR_ENTITY_TCR_GAMMA_VAR>",
|
2745 |
+
"single_word": false,
|
2746 |
+
"lstrip": false,
|
2747 |
+
"rstrip": false,
|
2748 |
+
"normalized": false,
|
2749 |
+
"special": true
|
2750 |
+
},
|
2751 |
+
{
|
2752 |
+
"id": 305,
|
2753 |
+
"content": "<SCALAR>",
|
2754 |
+
"single_word": false,
|
2755 |
+
"lstrip": false,
|
2756 |
+
"rstrip": false,
|
2757 |
+
"normalized": false,
|
2758 |
+
"special": true
|
2759 |
+
},
|
2760 |
+
{
|
2761 |
+
"id": 306,
|
2762 |
+
"content": "<VECTOR>",
|
2763 |
+
"single_word": false,
|
2764 |
+
"lstrip": false,
|
2765 |
+
"rstrip": false,
|
2766 |
+
"normalized": false,
|
2767 |
+
"special": true
|
2768 |
+
},
|
2769 |
+
{
|
2770 |
+
"id": 307,
|
2771 |
+
"content": "<MASKED_SCALAR>",
|
2772 |
+
"single_word": false,
|
2773 |
+
"lstrip": false,
|
2774 |
+
"rstrip": false,
|
2775 |
+
"normalized": false,
|
2776 |
+
"special": true
|
2777 |
+
},
|
2778 |
+
{
|
2779 |
+
"id": 308,
|
2780 |
+
"content": "<MASKED_VECTOR>",
|
2781 |
+
"single_word": false,
|
2782 |
+
"lstrip": false,
|
2783 |
+
"rstrip": false,
|
2784 |
+
"normalized": false,
|
2785 |
+
"special": true
|
2786 |
+
},
|
2787 |
+
{
|
2788 |
+
"id": 309,
|
2789 |
+
"content": "<AUTOENCODER_LATENT_LOG_VARIANCE>",
|
2790 |
+
"single_word": false,
|
2791 |
+
"lstrip": false,
|
2792 |
+
"rstrip": false,
|
2793 |
+
"normalized": false,
|
2794 |
+
"special": true
|
2795 |
+
},
|
2796 |
+
{
|
2797 |
+
"id": 310,
|
2798 |
+
"content": "<AUTOENCODER_LATENT_MEAN>",
|
2799 |
+
"single_word": false,
|
2800 |
+
"lstrip": false,
|
2801 |
+
"rstrip": false,
|
2802 |
+
"normalized": false,
|
2803 |
+
"special": true
|
2804 |
+
},
|
2805 |
+
{
|
2806 |
+
"id": 311,
|
2807 |
+
"content": "<AUTOENCODER_LATENT_SAMPLED_Z>",
|
2808 |
+
"single_word": false,
|
2809 |
+
"lstrip": false,
|
2810 |
+
"rstrip": false,
|
2811 |
+
"normalized": false,
|
2812 |
+
"special": true
|
2813 |
+
},
|
2814 |
+
{
|
2815 |
+
"id": 312,
|
2816 |
+
"content": "<AUTOENCODER_TASK>",
|
2817 |
+
"single_word": false,
|
2818 |
+
"lstrip": false,
|
2819 |
+
"rstrip": false,
|
2820 |
+
"normalized": false,
|
2821 |
+
"special": true
|
2822 |
+
},
|
2823 |
+
{
|
2824 |
+
"id": 313,
|
2825 |
+
"content": "<DECODED_FROM_LATENT>",
|
2826 |
+
"single_word": false,
|
2827 |
+
"lstrip": false,
|
2828 |
+
"rstrip": false,
|
2829 |
+
"normalized": false,
|
2830 |
+
"special": true
|
2831 |
+
},
|
2832 |
+
{
|
2833 |
+
"id": 314,
|
2834 |
+
"content": "<BBBP>",
|
2835 |
+
"single_word": false,
|
2836 |
+
"lstrip": false,
|
2837 |
+
"rstrip": false,
|
2838 |
+
"normalized": false,
|
2839 |
+
"special": true
|
2840 |
+
},
|
2841 |
+
{
|
2842 |
+
"id": 315,
|
2843 |
+
"content": "<FDA_APPR>",
|
2844 |
+
"single_word": false,
|
2845 |
+
"lstrip": false,
|
2846 |
+
"rstrip": false,
|
2847 |
+
"normalized": false,
|
2848 |
+
"special": true
|
2849 |
+
},
|
2850 |
+
{
|
2851 |
+
"id": 316,
|
2852 |
+
"content": "<HIV_ACTIVITY>",
|
2853 |
+
"single_word": false,
|
2854 |
+
"lstrip": false,
|
2855 |
+
"rstrip": false,
|
2856 |
+
"normalized": false,
|
2857 |
+
"special": true
|
2858 |
+
}
|
2859 |
+
],
|
2860 |
+
"normalizer": null,
|
2861 |
+
"pre_tokenizer": {
|
2862 |
+
"type": "Sequence",
|
2863 |
+
"pretokenizers": [
|
2864 |
+
{
|
2865 |
+
"type": "Split",
|
2866 |
+
"pattern": {
|
2867 |
+
"Regex": "<.*?>|\\[.*?\\]|\\S"
|
2868 |
+
},
|
2869 |
+
"behavior": "Removed",
|
2870 |
+
"invert": true
|
2871 |
+
}
|
2872 |
+
]
|
2873 |
+
},
|
2874 |
+
"post_processor": null,
|
2875 |
+
"decoder": null,
|
2876 |
+
"model": {
|
2877 |
+
"type": "WordLevel",
|
2878 |
+
"vocab": {
|
2879 |
+
"<UNK>": 0,
|
2880 |
+
"<PAD>": 1,
|
2881 |
+
"<CLS>": 2,
|
2882 |
+
"<SEP>": 3,
|
2883 |
+
"<MASK>": 4,
|
2884 |
+
"<EOS>": 5,
|
2885 |
+
"<MOLECULAR_ENTITY>": 6,
|
2886 |
+
"<GLOBAL_INTERACTION_ATTRIBUTES>": 7,
|
2887 |
+
"<MOLECULAR_ENTITY_ANTIGEN>": 8,
|
2888 |
+
"<MOLECULAR_ENTITY_EPITOPE>": 9,
|
2889 |
+
"<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN>": 10,
|
2890 |
+
"<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN>": 11,
|
2891 |
+
"<MOLECULAR_ENTITY_TCR_ALPHA_CHAIN>": 12,
|
2892 |
+
"<MOLECULAR_ENTITY_TCR_BETA_VDJ>": 13,
|
2893 |
+
"<MOLECULAR_ENTITY_TCR_BETA_CDR3>": 14,
|
2894 |
+
"<BINDING_AFFINITY_CLASS>": 15,
|
2895 |
+
"<DECODER_START>": 16,
|
2896 |
+
"<BINDING>": 17,
|
2897 |
+
"<FILLIN>": 18,
|
2898 |
+
"<REORDER>": 19,
|
2899 |
+
"<TOAA>": 20,
|
2900 |
+
"<ACTIVE>": 21,
|
2901 |
+
"<GENESEQ>": 22,
|
2902 |
+
"<INCREASE>": 23,
|
2903 |
+
"<DECREASE>": 24,
|
2904 |
+
"<STRUCTURE>": 25,
|
2905 |
+
"<DISTANCE>": 26,
|
2906 |
+
"<SOLUBILITY>": 27,
|
2907 |
+
"<TOXICITY>": 28,
|
2908 |
+
"<AB>": 29,
|
2909 |
+
"<ISACTIVE>": 30,
|
2910 |
+
"<ISSYNTHETIC>": 31,
|
2911 |
+
"<PENETR>": 32,
|
2912 |
+
"<ABSORPTION>": 33,
|
2913 |
+
"<DISTRIBUTION>": 34,
|
2914 |
+
"<METABOLISM>": 35,
|
2915 |
+
"<EXCRETION>": 36,
|
2916 |
+
"<FLUORESCENCE>": 37,
|
2917 |
+
"<STABILITY>": 38,
|
2918 |
+
"<DISORDER>": 39,
|
2919 |
+
"<DISEASE>": 40,
|
2920 |
+
"<BINARY>": 41,
|
2921 |
+
"<REGRESSION>": 42,
|
2922 |
+
"<ORGANISM>": 43,
|
2923 |
+
"<0>": 44,
|
2924 |
+
"<1>": 45,
|
2925 |
+
"<2>": 46,
|
2926 |
+
"<3>": 47,
|
2927 |
+
"<4>": 48,
|
2928 |
+
"<5>": 49,
|
2929 |
+
"<6>": 50,
|
2930 |
+
"<7>": 51,
|
2931 |
+
"<8>": 52,
|
2932 |
+
"<9>": 53,
|
2933 |
+
"<.>": 54,
|
2934 |
+
"<YES>": 55,
|
2935 |
+
"<NO>": 56,
|
2936 |
+
"<SENTINEL_ID_0>": 57,
|
2937 |
+
"<SENTINEL_ID_1>": 58,
|
2938 |
+
"<SENTINEL_ID_2>": 59,
|
2939 |
+
"<SENTINEL_ID_3>": 60,
|
2940 |
+
"<SENTINEL_ID_4>": 61,
|
2941 |
+
"<SENTINEL_ID_5>": 62,
|
2942 |
+
"<SENTINEL_ID_6>": 63,
|
2943 |
+
"<SENTINEL_ID_7>": 64,
|
2944 |
+
"<SENTINEL_ID_8>": 65,
|
2945 |
+
"<SENTINEL_ID_9>": 66,
|
2946 |
+
"<SENTINEL_ID_10>": 67,
|
2947 |
+
"<SENTINEL_ID_11>": 68,
|
2948 |
+
"<SENTINEL_ID_12>": 69,
|
2949 |
+
"<SENTINEL_ID_13>": 70,
|
2950 |
+
"<SENTINEL_ID_14>": 71,
|
2951 |
+
"<SENTINEL_ID_15>": 72,
|
2952 |
+
"<SENTINEL_ID_16>": 73,
|
2953 |
+
"<SENTINEL_ID_17>": 74,
|
2954 |
+
"<SENTINEL_ID_18>": 75,
|
2955 |
+
"<SENTINEL_ID_19>": 76,
|
2956 |
+
"<SENTINEL_ID_20>": 77,
|
2957 |
+
"<SENTINEL_ID_21>": 78,
|
2958 |
+
"<SENTINEL_ID_22>": 79,
|
2959 |
+
"<SENTINEL_ID_23>": 80,
|
2960 |
+
"<SENTINEL_ID_24>": 81,
|
2961 |
+
"<SENTINEL_ID_25>": 82,
|
2962 |
+
"<SENTINEL_ID_26>": 83,
|
2963 |
+
"<SENTINEL_ID_27>": 84,
|
2964 |
+
"<SENTINEL_ID_28>": 85,
|
2965 |
+
"<SENTINEL_ID_29>": 86,
|
2966 |
+
"<SENTINEL_ID_30>": 87,
|
2967 |
+
"<SENTINEL_ID_31>": 88,
|
2968 |
+
"<SENTINEL_ID_32>": 89,
|
2969 |
+
"<SENTINEL_ID_33>": 90,
|
2970 |
+
"<SENTINEL_ID_34>": 91,
|
2971 |
+
"<SENTINEL_ID_35>": 92,
|
2972 |
+
"<SENTINEL_ID_36>": 93,
|
2973 |
+
"<SENTINEL_ID_37>": 94,
|
2974 |
+
"<SENTINEL_ID_38>": 95,
|
2975 |
+
"<SENTINEL_ID_39>": 96,
|
2976 |
+
"<SENTINEL_ID_40>": 97,
|
2977 |
+
"<SENTINEL_ID_41>": 98,
|
2978 |
+
"<SENTINEL_ID_42>": 99,
|
2979 |
+
"<SENTINEL_ID_43>": 100,
|
2980 |
+
"<SENTINEL_ID_44>": 101,
|
2981 |
+
"<SENTINEL_ID_45>": 102,
|
2982 |
+
"<SENTINEL_ID_46>": 103,
|
2983 |
+
"<SENTINEL_ID_47>": 104,
|
2984 |
+
"<SENTINEL_ID_48>": 105,
|
2985 |
+
"<SENTINEL_ID_49>": 106,
|
2986 |
+
"<SENTINEL_ID_50>": 107,
|
2987 |
+
"<SENTINEL_ID_51>": 108,
|
2988 |
+
"<SENTINEL_ID_52>": 109,
|
2989 |
+
"<SENTINEL_ID_53>": 110,
|
2990 |
+
"<SENTINEL_ID_54>": 111,
|
2991 |
+
"<SENTINEL_ID_55>": 112,
|
2992 |
+
"<SENTINEL_ID_56>": 113,
|
2993 |
+
"<SENTINEL_ID_57>": 114,
|
2994 |
+
"<SENTINEL_ID_58>": 115,
|
2995 |
+
"<SENTINEL_ID_59>": 116,
|
2996 |
+
"<SENTINEL_ID_60>": 117,
|
2997 |
+
"<SENTINEL_ID_61>": 118,
|
2998 |
+
"<SENTINEL_ID_62>": 119,
|
2999 |
+
"<SENTINEL_ID_63>": 120,
|
3000 |
+
"<SENTINEL_ID_64>": 121,
|
3001 |
+
"<SENTINEL_ID_65>": 122,
|
3002 |
+
"<SENTINEL_ID_66>": 123,
|
3003 |
+
"<SENTINEL_ID_67>": 124,
|
3004 |
+
"<SENTINEL_ID_68>": 125,
|
3005 |
+
"<SENTINEL_ID_69>": 126,
|
3006 |
+
"<SENTINEL_ID_70>": 127,
|
3007 |
+
"<SENTINEL_ID_71>": 128,
|
3008 |
+
"<SENTINEL_ID_72>": 129,
|
3009 |
+
"<SENTINEL_ID_73>": 130,
|
3010 |
+
"<SENTINEL_ID_74>": 131,
|
3011 |
+
"<SENTINEL_ID_75>": 132,
|
3012 |
+
"<SENTINEL_ID_76>": 133,
|
3013 |
+
"<SENTINEL_ID_77>": 134,
|
3014 |
+
"<SENTINEL_ID_78>": 135,
|
3015 |
+
"<SENTINEL_ID_79>": 136,
|
3016 |
+
"<SENTINEL_ID_80>": 137,
|
3017 |
+
"<SENTINEL_ID_81>": 138,
|
3018 |
+
"<SENTINEL_ID_82>": 139,
|
3019 |
+
"<SENTINEL_ID_83>": 140,
|
3020 |
+
"<SENTINEL_ID_84>": 141,
|
3021 |
+
"<SENTINEL_ID_85>": 142,
|
3022 |
+
"<SENTINEL_ID_86>": 143,
|
3023 |
+
"<SENTINEL_ID_87>": 144,
|
3024 |
+
"<SENTINEL_ID_88>": 145,
|
3025 |
+
"<SENTINEL_ID_89>": 146,
|
3026 |
+
"<SENTINEL_ID_90>": 147,
|
3027 |
+
"<SENTINEL_ID_91>": 148,
|
3028 |
+
"<SENTINEL_ID_92>": 149,
|
3029 |
+
"<SENTINEL_ID_93>": 150,
|
3030 |
+
"<SENTINEL_ID_94>": 151,
|
3031 |
+
"<SENTINEL_ID_95>": 152,
|
3032 |
+
"<SENTINEL_ID_96>": 153,
|
3033 |
+
"<SENTINEL_ID_97>": 154,
|
3034 |
+
"<SENTINEL_ID_98>": 155,
|
3035 |
+
"<SENTINEL_ID_99>": 156,
|
3036 |
+
"<SENTINEL_ID_100>": 157,
|
3037 |
+
"<SENTINEL_ID_101>": 158,
|
3038 |
+
"<SENTINEL_ID_102>": 159,
|
3039 |
+
"<SENTINEL_ID_103>": 160,
|
3040 |
+
"<SENTINEL_ID_104>": 161,
|
3041 |
+
"<SENTINEL_ID_105>": 162,
|
3042 |
+
"<SENTINEL_ID_106>": 163,
|
3043 |
+
"<SENTINEL_ID_107>": 164,
|
3044 |
+
"<SENTINEL_ID_108>": 165,
|
3045 |
+
"<SENTINEL_ID_109>": 166,
|
3046 |
+
"<SENTINEL_ID_110>": 167,
|
3047 |
+
"<SENTINEL_ID_111>": 168,
|
3048 |
+
"<SENTINEL_ID_112>": 169,
|
3049 |
+
"<SENTINEL_ID_113>": 170,
|
3050 |
+
"<SENTINEL_ID_114>": 171,
|
3051 |
+
"<SENTINEL_ID_115>": 172,
|
3052 |
+
"<SENTINEL_ID_116>": 173,
|
3053 |
+
"<SENTINEL_ID_117>": 174,
|
3054 |
+
"<SENTINEL_ID_118>": 175,
|
3055 |
+
"<SENTINEL_ID_119>": 176,
|
3056 |
+
"<SENTINEL_ID_120>": 177,
|
3057 |
+
"<SENTINEL_ID_121>": 178,
|
3058 |
+
"<SENTINEL_ID_122>": 179,
|
3059 |
+
"<SENTINEL_ID_123>": 180,
|
3060 |
+
"<SENTINEL_ID_124>": 181,
|
3061 |
+
"<SENTINEL_ID_125>": 182,
|
3062 |
+
"<SENTINEL_ID_126>": 183,
|
3063 |
+
"<SENTINEL_ID_127>": 184,
|
3064 |
+
"<SENTINEL_ID_128>": 185,
|
3065 |
+
"<SENTINEL_ID_129>": 186,
|
3066 |
+
"<SENTINEL_ID_130>": 187,
|
3067 |
+
"<SENTINEL_ID_131>": 188,
|
3068 |
+
"<SENTINEL_ID_132>": 189,
|
3069 |
+
"<SENTINEL_ID_133>": 190,
|
3070 |
+
"<SENTINEL_ID_134>": 191,
|
3071 |
+
"<SENTINEL_ID_135>": 192,
|
3072 |
+
"<SENTINEL_ID_136>": 193,
|
3073 |
+
"<SENTINEL_ID_137>": 194,
|
3074 |
+
"<SENTINEL_ID_138>": 195,
|
3075 |
+
"<SENTINEL_ID_139>": 196,
|
3076 |
+
"<SENTINEL_ID_140>": 197,
|
3077 |
+
"<SENTINEL_ID_141>": 198,
|
3078 |
+
"<SENTINEL_ID_142>": 199,
|
3079 |
+
"<SENTINEL_ID_143>": 200,
|
3080 |
+
"<SENTINEL_ID_144>": 201,
|
3081 |
+
"<SENTINEL_ID_145>": 202,
|
3082 |
+
"<SENTINEL_ID_146>": 203,
|
3083 |
+
"<SENTINEL_ID_147>": 204,
|
3084 |
+
"<SENTINEL_ID_148>": 205,
|
3085 |
+
"<SENTINEL_ID_149>": 206,
|
3086 |
+
"<SENTINEL_ID_150>": 207,
|
3087 |
+
"<SENTINEL_ID_151>": 208,
|
3088 |
+
"<SENTINEL_ID_152>": 209,
|
3089 |
+
"<SENTINEL_ID_153>": 210,
|
3090 |
+
"<SENTINEL_ID_154>": 211,
|
3091 |
+
"<SENTINEL_ID_155>": 212,
|
3092 |
+
"<SENTINEL_ID_156>": 213,
|
3093 |
+
"<SENTINEL_ID_157>": 214,
|
3094 |
+
"<SENTINEL_ID_158>": 215,
|
3095 |
+
"<SENTINEL_ID_159>": 216,
|
3096 |
+
"<SENTINEL_ID_160>": 217,
|
3097 |
+
"<SENTINEL_ID_161>": 218,
|
3098 |
+
"<SENTINEL_ID_162>": 219,
|
3099 |
+
"<SENTINEL_ID_163>": 220,
|
3100 |
+
"<SENTINEL_ID_164>": 221,
|
3101 |
+
"<SENTINEL_ID_165>": 222,
|
3102 |
+
"<SENTINEL_ID_166>": 223,
|
3103 |
+
"<SENTINEL_ID_167>": 224,
|
3104 |
+
"<SENTINEL_ID_168>": 225,
|
3105 |
+
"<SENTINEL_ID_169>": 226,
|
3106 |
+
"<SENTINEL_ID_170>": 227,
|
3107 |
+
"<SENTINEL_ID_171>": 228,
|
3108 |
+
"<SENTINEL_ID_172>": 229,
|
3109 |
+
"<SENTINEL_ID_173>": 230,
|
3110 |
+
"<SENTINEL_ID_174>": 231,
|
3111 |
+
"<SENTINEL_ID_175>": 232,
|
3112 |
+
"<SENTINEL_ID_176>": 233,
|
3113 |
+
"<SENTINEL_ID_177>": 234,
|
3114 |
+
"<SENTINEL_ID_178>": 235,
|
3115 |
+
"<SENTINEL_ID_179>": 236,
|
3116 |
+
"<SENTINEL_ID_180>": 237,
|
3117 |
+
"<SENTINEL_ID_181>": 238,
|
3118 |
+
"<SENTINEL_ID_182>": 239,
|
3119 |
+
"<SENTINEL_ID_183>": 240,
|
3120 |
+
"<SENTINEL_ID_184>": 241,
|
3121 |
+
"<SENTINEL_ID_185>": 242,
|
3122 |
+
"<SENTINEL_ID_186>": 243,
|
3123 |
+
"<SENTINEL_ID_187>": 244,
|
3124 |
+
"<SENTINEL_ID_188>": 245,
|
3125 |
+
"<SENTINEL_ID_189>": 246,
|
3126 |
+
"<SENTINEL_ID_190>": 247,
|
3127 |
+
"<SENTINEL_ID_191>": 248,
|
3128 |
+
"<SENTINEL_ID_192>": 249,
|
3129 |
+
"<SENTINEL_ID_193>": 250,
|
3130 |
+
"<SENTINEL_ID_194>": 251,
|
3131 |
+
"<SENTINEL_ID_195>": 252,
|
3132 |
+
"<SENTINEL_ID_196>": 253,
|
3133 |
+
"<SENTINEL_ID_197>": 254,
|
3134 |
+
"<SENTINEL_ID_198>": 255,
|
3135 |
+
"<SENTINEL_ID_199>": 256,
|
3136 |
+
"<MOLECULAR_ENTITY_TYPE_ANTIGEN>": 257,
|
3137 |
+
"<MOLECULAR_ENTITY_TYPE_ANTIBODY_LIGHT_CHAIN>": 258,
|
3138 |
+
"<MOLECULAR_ENTITY_TYPE_ANTIBODY_HEAVY_CHAIN>": 259,
|
3139 |
+
"<ATTRIBUTE_ORGANISM>": 260,
|
3140 |
+
"<ATTRIBUTE_ORGANISM_HUMAN>": 261,
|
3141 |
+
"<ATTRIBUTE_ORGANISM_RABBIT>": 262,
|
3142 |
+
"<ATTRIBUTE_ORGANISM_RAT>": 263,
|
3143 |
+
"<ATTRIBUTE_ORGANISM_MOUSE>": 264,
|
3144 |
+
"<ATTRIBUTE_ORGANISM_MONKEY>": 265,
|
3145 |
+
"<ATTRIBUTE_ORGANISM_CAMEL>": 266,
|
3146 |
+
"<EPITOPE_PARATOPE_PREDICTION>": 267,
|
3147 |
+
"<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN_CDR1>": 268,
|
3148 |
+
"<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN_CDR3>": 269,
|
3149 |
+
"<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN_CDR3>": 270,
|
3150 |
+
"<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN_CDR2>": 271,
|
3151 |
+
"<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN_CDR2>": 272,
|
3152 |
+
"<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN_CDR1>": 273,
|
3153 |
+
"<MOLECULAR_ENTITY_GENERAL_PROTEIN>": 274,
|
3154 |
+
"<TIMESTEP>": 275,
|
3155 |
+
"<DIFFUSION>": 276,
|
3156 |
+
"<SEQUENCE_NATURAL_END>": 277,
|
3157 |
+
"<SMILES_SEQUENCE>": 278,
|
3158 |
+
"<SELFIES_SEQUENCE>": 279,
|
3159 |
+
"<AMINO_ACID_SEQUENCE>": 280,
|
3160 |
+
"<GENERAL_AFFINITY_CLASS>": 281,
|
3161 |
+
"<BACKSPACE>": 282,
|
3162 |
+
"<SEQUENCE_NATURAL_START>": 283,
|
3163 |
+
"<NOOP>": 284,
|
3164 |
+
"<TARGETED_ANTIBODY_DESIGN_ENCODER_ONLY_MODE>": 285,
|
3165 |
+
"<MOLECULAR_ENTITY_SMALL_MOLECULE>": 286,
|
3166 |
+
"<MOLECULAR_ENTITY_CELL_GENE_EXPRESSION_RANKED>": 287,
|
3167 |
+
"<CELL_TYPE_CLASS>": 288,
|
3168 |
+
"<TISSUE_TYPE_CLASS>": 289,
|
3169 |
+
"<CORRUPTED_AREA_START>": 290,
|
3170 |
+
"<CORRUPTED_AREA_END>": 291,
|
3171 |
+
"<MOLECULAR_ENTITY_MUTATED_PROTEIN_CHAIN>": 292,
|
3172 |
+
"<MOLECULAR_ENTITY_PROTEIN_CHAIN>": 293,
|
3173 |
+
"<COMPLEX_ENTITY>": 294,
|
3174 |
+
"<ALTERNATIVE>": 295,
|
3175 |
+
"<CDR3_REGION>": 296,
|
3176 |
+
"<GENERAL_CHAIN>": 297,
|
3177 |
+
"<SUBMOLECULAR_ENTITY>": 298,
|
3178 |
+
"<MUTATED>": 299,
|
3179 |
+
"<MOLECULAR_ENTITY_TCR_ALPHA_CDR3>": 300,
|
3180 |
+
"<MOLECULAR_ENTITY_TCR_DELTA_CDR3>": 301,
|
3181 |
+
"<MOLECULAR_ENTITY_TCR_DELTA_VAR>": 302,
|
3182 |
+
"<MOLECULAR_ENTITY_TCR_GAMMA_CDR3>": 303,
|
3183 |
+
"<MOLECULAR_ENTITY_TCR_GAMMA_VAR>": 304,
|
3184 |
+
"<SCALAR>": 305,
|
3185 |
+
"<VECTOR>": 306,
|
3186 |
+
"<MASKED_SCALAR>": 307,
|
3187 |
+
"<MASKED_VECTOR>": 308,
|
3188 |
+
"<AUTOENCODER_LATENT_LOG_VARIANCE>": 309,
|
3189 |
+
"<AUTOENCODER_LATENT_MEAN>": 310,
|
3190 |
+
"<AUTOENCODER_LATENT_SAMPLED_Z>": 311,
|
3191 |
+
"<AUTOENCODER_TASK>": 312,
|
3192 |
+
"<DECODED_FROM_LATENT>": 313,
|
3193 |
+
"<BBBP>": 314,
|
3194 |
+
"<FDA_APPR>": 315,
|
3195 |
+
"<HIV_ACTIVITY>": 316,
|
3196 |
+
"[CL:0000499]": 3522,
|
3197 |
+
"[CL:2000060]": 3523,
|
3198 |
+
"[CL:0000235]": 3524,
|
3199 |
+
"[CL:0002343]": 3525,
|
3200 |
+
"[CL:0000084]": 3526,
|
3201 |
+
"[CL:2000042]": 3527,
|
3202 |
+
"[CL:0000066]": 3528,
|
3203 |
+
"[CL:0008036]": 3529,
|
3204 |
+
"[CL:3000001]": 3530,
|
3205 |
+
"[CL:0000525]": 3531,
|
3206 |
+
"[CL:0000623]": 3532,
|
3207 |
+
"[CL:0009095]": 3533,
|
3208 |
+
"[CL:0000003]": 3534,
|
3209 |
+
"[CL:0001078]": 3535,
|
3210 |
+
"[CL:0000815]": 3536,
|
3211 |
+
"[CL:0002601]": 3537,
|
3212 |
+
"[CL:0000451]": 3538,
|
3213 |
+
"[CL:0009092]": 3539,
|
3214 |
+
"[CL:0002138]": 3540,
|
3215 |
+
"[CL:0000236]": 3541,
|
3216 |
+
"[CL:0000786]": 3542,
|
3217 |
+
"[CL:0000094]": 3543,
|
3218 |
+
"[CL:0002488]": 3544,
|
3219 |
+
"[CL:0000625]": 3545,
|
3220 |
+
"[CL:0000913]": 3546,
|
3221 |
+
"[CL:0000624]": 3547,
|
3222 |
+
"[CL:0000895]": 3548,
|
3223 |
+
"[CL:0000905]": 3549,
|
3224 |
+
"[CL:0000792]": 3550,
|
3225 |
+
"[CL:0000940]": 3551,
|
3226 |
+
"[CL:0000788]": 3552,
|
3227 |
+
"[CL:0000900]": 3553,
|
3228 |
+
"[CL:0000798]": 3554,
|
3229 |
+
"[CL:0000970]": 3555,
|
3230 |
+
"[CL:0000985]": 3556,
|
3231 |
+
"[CL:0000972]": 3557,
|
3232 |
+
"[CL:0000987]": 3558,
|
3233 |
+
"[CL:0011026]": 3559,
|
3234 |
+
"[CL:4023040]": 3560,
|
3235 |
+
"[CL:0002555]": 3561,
|
3236 |
+
"[CL:0000071]": 3562,
|
3237 |
+
"[CL:0002326]": 3563,
|
3238 |
+
"[CL:0008034]": 3564,
|
3239 |
+
"[CL:0002324]": 3565,
|
3240 |
+
"[CL:0000128]": 3566,
|
3241 |
+
"[CL:4023016]": 3567,
|
3242 |
+
"[CL:4023018]": 3568,
|
3243 |
+
"[CL:0000988]": 3569,
|
3244 |
+
"[CL:0002605]": 3570,
|
3245 |
+
"[CL:4023011]": 3571,
|
3246 |
+
"[CL:4023017]": 3572,
|
3247 |
+
"[CL:0000129]": 3573,
|
3248 |
+
"[CL:0002453]": 3574,
|
3249 |
+
"[CL:4023015]": 3575,
|
3250 |
+
"[CL:4023012]": 3576,
|
3251 |
+
"[CL:4023013]": 3577,
|
3252 |
+
"[CL:4023038]": 3578,
|
3253 |
+
"[CL:4023036]": 3579,
|
3254 |
+
"[CL:4023070]": 3580,
|
3255 |
+
"[CL:4023051]": 3581,
|
3256 |
+
"[CL:4023041]": 3582,
|
3257 |
+
"[CL:1001602]": 3583,
|
3258 |
+
"[CL:0000151]": 3584,
|
3259 |
+
"[CL:0000860]": 3585,
|
3260 |
+
"[CL:0000186]": 3586,
|
3261 |
+
"[CL:0002306]": 3587,
|
3262 |
+
"[CL:0000814]": 3588,
|
3263 |
+
"[CL:1001431]": 3589,
|
3264 |
+
"[CL:1001106]": 3590,
|
3265 |
+
"[CL:0000057]": 3591,
|
3266 |
+
"[CL:1000692]": 3592,
|
3267 |
+
"[CL:0000115]": 3593,
|
3268 |
+
"[CL:0000192]": 3594,
|
3269 |
+
"[CL:0002117]": 3595,
|
3270 |
+
"[CL:0000233]": 3596,
|
3271 |
+
"[CL:0001044]": 3597,
|
3272 |
+
"[CL:0001050]": 3598,
|
3273 |
+
"[CL:1001111]": 3599,
|
3274 |
+
"[CL:0000067]": 3600,
|
3275 |
+
"[CL:1001107]": 3601,
|
3276 |
+
"[CL:1001432]": 3602,
|
3277 |
+
"[CL:1000849]": 3603,
|
3278 |
+
"[CL:0000669]": 3604,
|
3279 |
+
"[CL:0000738]": 3605,
|
3280 |
+
"[CL:0000492]": 3606,
|
3281 |
+
"[CL:1000768]": 3607,
|
3282 |
+
"[CL:0000979]": 3608,
|
3283 |
+
"[CL:0000875]": 3609,
|
3284 |
+
"[CL:0000097]": 3610,
|
3285 |
+
"[CL:0000081]": 3611,
|
3286 |
+
"[CL:0000232]": 3612,
|
3287 |
+
"[CL:1001318]": 3613,
|
3288 |
+
"[CL:0002393]": 3614,
|
3289 |
+
"[CL:0000653]": 3615,
|
3290 |
+
"[CL:1000597]": 3616,
|
3291 |
+
"[CL:1000452]": 3617,
|
3292 |
+
"[CL:0002319]": 3618,
|
3293 |
+
"[CL:0019026]": 3619,
|
3294 |
+
"[CL:0019029]": 3620,
|
3295 |
+
"[CL:0019028]": 3621,
|
3296 |
+
"[CL:0019022]": 3622,
|
3297 |
+
"[CL:0010011]": 3623,
|
3298 |
+
"[CL:0000091]": 3624,
|
3299 |
+
"[CL:0000863]": 3625,
|
3300 |
+
"[CL:0000679]": 3626,
|
3301 |
+
"[CL:0000632]": 3627,
|
3302 |
+
"[CL:0013000]": 3628,
|
3303 |
+
"[CL:0001054]": 3629,
|
3304 |
+
"[CL:1000488]": 3630,
|
3305 |
+
"[CL:0000939]": 3631,
|
3306 |
+
"[CL:0000789]": 3632,
|
3307 |
+
"[CL:0000182]": 3633,
|
3308 |
+
"[CL:0019021]": 3634,
|
3309 |
+
"[CL:0000904]": 3635,
|
3310 |
+
"[CL:0001042]": 3636,
|
3311 |
+
"[CL:0000785]": 3637,
|
3312 |
+
"[CL:0002038]": 3638,
|
3313 |
+
"[CL:0000938]": 3639,
|
3314 |
+
"[CL:0000980]": 3640,
|
3315 |
+
"[CL:0000816]": 3641,
|
3316 |
+
"[CL:0000784]": 3642,
|
3317 |
+
"[CL:0002396]": 3643,
|
3318 |
+
"[CL:0000764]": 3644,
|
3319 |
+
"[CL:0000359]": 3645,
|
3320 |
+
"[CL:0000782]": 3646,
|
3321 |
+
"[CL:0001056]": 3647,
|
3322 |
+
"[CL:0000695]": 3648,
|
3323 |
+
"[CL:0000986]": 3649,
|
3324 |
+
"[CL:0001024]": 3650,
|
3325 |
+
"[CL:0000576]": 3651,
|
3326 |
+
"[CL:0001077]": 3652,
|
3327 |
+
"[CL:0008001]": 3653,
|
3328 |
+
"[CL:0000545]": 3654,
|
3329 |
+
"[CL:0000546]": 3655,
|
3330 |
+
"[CL:0001081]": 3656,
|
3331 |
+
"[CL:0001066]": 3657,
|
3332 |
+
"[CL:0000990]": 3658,
|
3333 |
+
"[CL:0000542]": 3659,
|
3334 |
+
"[CL:0000604]": 3660,
|
3335 |
+
"[CL:0000748]": 3661,
|
3336 |
+
"[CL:0000125]": 3662,
|
3337 |
+
"[CL:0000573]": 3663,
|
3338 |
+
"[CL:0000561]": 3664,
|
3339 |
+
"[CL:0000745]": 3665,
|
3340 |
+
"[CL:0000749]": 3666,
|
3341 |
+
"[CL:0000750]": 3667,
|
3342 |
+
"[CL:0000636]": 3668,
|
3343 |
+
"[CL:0000740]": 3669,
|
3344 |
+
"[CL:0000127]": 3670,
|
3345 |
+
"[CL:0002586]": 3671,
|
3346 |
+
"[CL:0000583]": 3672,
|
3347 |
+
"[CL:0002063]": 3673,
|
3348 |
+
"[CL:0002062]": 3674,
|
3349 |
+
"[CL:0000158]": 3675,
|
3350 |
+
"[CL:0000064]": 3676,
|
3351 |
+
"[CL:0000138]": 3677,
|
3352 |
+
"[CL:0000160]": 3678,
|
3353 |
+
"[CL:0000646]": 3679,
|
3354 |
+
"[CL:0000595]": 3680,
|
3355 |
+
"[CL:1000413]": 3681,
|
3356 |
+
"[CL:0002543]": 3682,
|
3357 |
+
"[CL:1000223]": 3683,
|
3358 |
+
"[CL:0002341]": 3684,
|
3359 |
+
"[CL:1000296]": 3685,
|
3360 |
+
"[CL:0002340]": 3686,
|
3361 |
+
"[CL:1000304]": 3687,
|
3362 |
+
"[CL:0002399]": 3688,
|
3363 |
+
"[CL:1000487]": 3689,
|
3364 |
+
"[CL:0000810]": 3690,
|
3365 |
+
"[CL:1000305]": 3691,
|
3366 |
+
"[CL:2000059]": 3692,
|
3367 |
+
"[CL:0002622]": 3693,
|
3368 |
+
"[CL:1000299]": 3694,
|
3369 |
+
"[CL:0000811]": 3695,
|
3370 |
+
"[CL:0002394]": 3696,
|
3371 |
+
"[CL:0000165]": 3697,
|
3372 |
+
"[CL:0000775]": 3698,
|
3373 |
+
"[CL:0000908]": 3699,
|
3374 |
+
"[CL:0000921]": 3700,
|
3375 |
+
"[CL:0000897]": 3701,
|
3376 |
+
"[CL:1000398]": 3702,
|
3377 |
+
"[CL:0001065]": 3703,
|
3378 |
+
"[CL:0000787]": 3704,
|
3379 |
+
"[CL:1000329]": 3705,
|
3380 |
+
"[CL:0000319]": 3706,
|
3381 |
+
"[CL:0000037]": 3707,
|
3382 |
+
"[CL:1000330]": 3708,
|
3383 |
+
"[CL:2000055]": 3709,
|
3384 |
+
"[CL:0005006]": 3710,
|
3385 |
+
"[CL:0002538]": 3711,
|
3386 |
+
"[CL:0000767]": 3712,
|
3387 |
+
"[CL:0000809]": 3713,
|
3388 |
+
"[CL:0002623]": 3714,
|
3389 |
+
"[CL:0000746]": 3715,
|
3390 |
+
"[CL:1000432]": 3716,
|
3391 |
+
"[CL:0002071]": 3717,
|
3392 |
+
"[CL:0010008]": 3718,
|
3393 |
+
"[CL:0000909]": 3719,
|
3394 |
+
"[CL:4006000]": 3720,
|
3395 |
+
"[CL:0000763]": 3721,
|
3396 |
+
"[CL:0000068]": 3722,
|
3397 |
+
"[CL:0000575]": 3723,
|
3398 |
+
"[CL:1001428]": 3724,
|
3399 |
+
"[CL:0002064]": 3725,
|
3400 |
+
"[CL:0000287]": 3726,
|
3401 |
+
"[CL:0009009]": 3727,
|
3402 |
+
"[CL:0002079]": 3728,
|
3403 |
+
"[CL:0000134]": 3729,
|
3404 |
+
"[CL:0002677]": 3730,
|
3405 |
+
"[CL:0002363]": 3731,
|
3406 |
+
"[CL:0000794]": 3732,
|
3407 |
+
"[CL:0002585]": 3733,
|
3408 |
+
"[CL:0002503]": 3734,
|
3409 |
+
"[CL:0000038]": 3735,
|
3410 |
+
"[CL:1000320]": 3736,
|
3411 |
+
"[CL:0002144]": 3737,
|
3412 |
+
"[CL:0002410]": 3738,
|
3413 |
+
"[CL:0008011]": 3739,
|
3414 |
+
"[CL:0002139]": 3740,
|
3415 |
+
"[CL:0002548]": 3741,
|
3416 |
+
"[CL:0000148]": 3742,
|
3417 |
+
"[CL:0002366]": 3743,
|
3418 |
+
"[CL:0000185]": 3744,
|
3419 |
+
"[CL:0000681]": 3745,
|
3420 |
+
"[CL:0002149]": 3746,
|
3421 |
+
"[CL:0000187]": 3747,
|
3422 |
+
"[CL:0000034]": 3748,
|
3423 |
+
"[CL:0009016]": 3749,
|
3424 |
+
"[CL:0002365]": 3750,
|
3425 |
+
"[CL:0000190]": 3751,
|
3426 |
+
"[CL:0000114]": 3752,
|
3427 |
+
"[CL:1000436]": 3753,
|
3428 |
+
"[CL:0009011]": 3754,
|
3429 |
+
"[CL:0009005]": 3755,
|
3430 |
+
"[CL:0000136]": 3756,
|
3431 |
+
"[CL:0000189]": 3757,
|
3432 |
+
"[CL:0000131]": 3758,
|
3433 |
+
"[CL:0000049]": 3759,
|
3434 |
+
"[CL:1001516]": 3760,
|
3435 |
+
"[CL:0000312]": 3761,
|
3436 |
+
"[CL:0000808]": 3762,
|
3437 |
+
"[CL:0000453]": 3763,
|
3438 |
+
"[CL:0000169]": 3764,
|
3439 |
+
"[CL:0002303]": 3765,
|
3440 |
+
"[CL:0019032]": 3766,
|
3441 |
+
"[CL:0000841]": 3767,
|
3442 |
+
"[CL:0002673]": 3768,
|
3443 |
+
"[CL:0000388]": 3769,
|
3444 |
+
"[CL:0002370]": 3770,
|
3445 |
+
"[CL:0000077]": 3771,
|
3446 |
+
"[CL:0002518]": 3772,
|
3447 |
+
"[CL:0000807]": 3773,
|
3448 |
+
"[CL:0000584]": 3774,
|
3449 |
+
"[CL:1000334]": 3775,
|
3450 |
+
"[CL:0000894]": 3776,
|
3451 |
+
"[CL:1000271]": 3777,
|
3452 |
+
"[CL:2000016]": 3778,
|
3453 |
+
"[CL:1000495]": 3779,
|
3454 |
+
"[CL:1000343]": 3780,
|
3455 |
+
"[CL:0002250]": 3781,
|
3456 |
+
"[CL:0009012]": 3782,
|
3457 |
+
"[CL:0002598]": 3783,
|
3458 |
+
"[CL:0009017]": 3784,
|
3459 |
+
"[CL:0017000]": 3785,
|
3460 |
+
"[CL:1000331]": 3786,
|
3461 |
+
"[CL:0000019]": 3787,
|
3462 |
+
"[CL:0002573]": 3788,
|
3463 |
+
"[CL:0000188]": 3789,
|
3464 |
+
"[CL:0000893]": 3790,
|
3465 |
+
"[CL:0002320]": 3791,
|
3466 |
+
"[CL:0000171]": 3792,
|
3467 |
+
"[CL:0002275]": 3793,
|
3468 |
+
"[CL:1001589]": 3794,
|
3469 |
+
"[CL:0000823]": 3795,
|
3470 |
+
"[CL:0000173]": 3796,
|
3471 |
+
"[CL:0001058]": 3797,
|
3472 |
+
"[CL:0000934]": 3798,
|
3473 |
+
"[CL:0000818]": 3799,
|
3474 |
+
"[CL:0001043]": 3800,
|
3475 |
+
"[CL:0001049]": 3801,
|
3476 |
+
"[CL:0000776]": 3802,
|
3477 |
+
"[CL:0005012]": 3803,
|
3478 |
+
"[CL:0000982]": 3804,
|
3479 |
+
"[CL:0000984]": 3805,
|
3480 |
+
"[CL:0001076]": 3806,
|
3481 |
+
"[CL:0002489]": 3807,
|
3482 |
+
"[CL:1000272]": 3808,
|
3483 |
+
"[CL:0019019]": 3809,
|
3484 |
+
"[CL:0000135]": 3810,
|
3485 |
+
"[CL:0002057]": 3811,
|
3486 |
+
"[CL:0000791]": 3812,
|
3487 |
+
"[CL:0002397]": 3813,
|
3488 |
+
"[CL:0000556]": 3814,
|
3489 |
+
"[CL:0000800]": 3815,
|
3490 |
+
"[CL:0000076]": 3816,
|
3491 |
+
"[CL:0000121]": 3817,
|
3492 |
+
"[CL:0000498]": 3818,
|
3493 |
+
"[CL:0000120]": 3819,
|
3494 |
+
"[CL:0000123]": 3820,
|
3495 |
+
"[CL:0000122]": 3821,
|
3496 |
+
"[CL:0000555]": 3822,
|
3497 |
+
"[CL:0000126]": 3823,
|
3498 |
+
"[CL:0000540]": 3824,
|
3499 |
+
"[CL:0000164]": 3825,
|
3500 |
+
"[CL:0002425]": 3826,
|
3501 |
+
"[CL:0000817]": 3827,
|
3502 |
+
"[CL:0000557]": 3828,
|
3503 |
+
"[CL:0000826]": 3829,
|
3504 |
+
"[CL:0000050]": 3830,
|
3505 |
+
"[CL:0000834]": 3831,
|
3506 |
+
"[CL:0002045]": 3832,
|
3507 |
+
"[CL:0001082]": 3833,
|
3508 |
+
"[CL:0000617]": 3834,
|
3509 |
+
"[CL:0002563]": 3835,
|
3510 |
+
"[CL:0002350]": 3836,
|
3511 |
+
"[CL:1000309]": 3837,
|
3512 |
+
"[CL:2000041]": 3838,
|
3513 |
+
"[CL:0000166]": 3839,
|
3514 |
+
"[CL:0005025]": 3840,
|
3515 |
+
"[CL:0000765]": 3841,
|
3516 |
+
"[CL:2000046]": 3842,
|
3517 |
+
"[CL:0002129]": 3843,
|
3518 |
+
"[CL:0008019]": 3844,
|
3519 |
+
"[CL:0000216]": 3845,
|
3520 |
+
"[CL:0000501]": 3846,
|
3521 |
+
"[CL:0002094]": 3847,
|
3522 |
+
"[CL:0000630]": 3848,
|
3523 |
+
"[CL:2000064]": 3849,
|
3524 |
+
"[CL:0000178]": 3850,
|
3525 |
+
"[CL:0000586]": 3851,
|
3526 |
+
"[CL:0000024]": 3852,
|
3527 |
+
"[CL:0002371]": 3853,
|
3528 |
+
"[CL:0000670]": 3854,
|
3529 |
+
"[CL:0000023]": 3855,
|
3530 |
+
"[CL:0000015]": 3856,
|
3531 |
+
"[CL:0000650]": 3857,
|
3532 |
+
"[CL:0005026]": 3858,
|
3533 |
+
"[CL:0000209]": 3859,
|
3534 |
+
"[CL:0002632]": 3860,
|
3535 |
+
"[CL:0000622]": 3861,
|
3536 |
+
"[CL:0000594]": 3862,
|
3537 |
+
"[CL:0000163]": 3863,
|
3538 |
+
"[CL:0007011]": 3864,
|
3539 |
+
"[CL:0011101]": 3865,
|
3540 |
+
"[CL:0002097]": 3866,
|
3541 |
+
"[CL:0000210]": 3867,
|
3542 |
+
"[CL:0011103]": 3868,
|
3543 |
+
"[CL:0002204]": 3869,
|
3544 |
+
"[CL:0000145]": 3870,
|
3545 |
+
"[CL:0000397]": 3871,
|
3546 |
+
"[CL:0000103]": 3872,
|
3547 |
+
"[CL:0002293]": 3873,
|
3548 |
+
"[CL:0011004]": 3874,
|
3549 |
+
"[CL:0002504]": 3875,
|
3550 |
+
"[CL:0000677]": 3876,
|
3551 |
+
"[CL:1000279]": 3877,
|
3552 |
+
"[CL:0002088]": 3878,
|
3553 |
+
"[CL:0000898]": 3879,
|
3554 |
+
"[CL:0019031]": 3880,
|
3555 |
+
"[CL:0008015]": 3881,
|
3556 |
+
"[CL:0000100]": 3882,
|
3557 |
+
"[CL:1000275]": 3883,
|
3558 |
+
"[CL:0000099]": 3884,
|
3559 |
+
"[CL:0001071]": 3885,
|
3560 |
+
"[CL:0000766]": 3886,
|
3561 |
+
"[CL:0000082]": 3887,
|
3562 |
+
"[CL:2000032]": 3888,
|
3563 |
+
"[CL:0011115]": 3889,
|
3564 |
+
"[CL:0000682]": 3890,
|
3565 |
+
"[CL:0000906]": 3891,
|
3566 |
+
"[CL:0000896]": 3892,
|
3567 |
+
"[CL:0002131]": 3893,
|
3568 |
+
"[CL:0000813]": 3894,
|
3569 |
+
"[CL:0000079]": 3895,
|
3570 |
+
"[CL:2000095]": 3896,
|
3571 |
+
"[CL:1000909]": 3897,
|
3572 |
+
"[CL:0005010]": 3898,
|
3573 |
+
"[CL:0002322]": 3899,
|
3574 |
+
"[CL:0000890]": 3900,
|
3575 |
+
"[CL:0002258]": 3901,
|
3576 |
+
"[CL:0001063]": 3902,
|
3577 |
+
"[CL:0000696]": 3903,
|
3578 |
+
"[CL:0005019]": 3904,
|
3579 |
+
"[CL:0005022]": 3905,
|
3580 |
+
"[CL:0002627]": 3906,
|
3581 |
+
"[CL:0002629]": 3907,
|
3582 |
+
"[CL:0005011]": 3908,
|
3583 |
+
"[CL:0005009]": 3909,
|
3584 |
+
"[CL:1000892]": 3910,
|
3585 |
+
"[CL:0002201]": 3911,
|
3586 |
+
"[CL:0000790]": 3912,
|
3587 |
+
"[CL:1001433]": 3913,
|
3588 |
+
"[CL:0000155]": 3914,
|
3589 |
+
"[CL:0002633]": 3915,
|
3590 |
+
"[CL:1000143]": 3916,
|
3591 |
+
"[CL:0001057]": 3917,
|
3592 |
+
"[CL:0009002]": 3918,
|
3593 |
+
"[CL:0002241]": 3919,
|
3594 |
+
"[CL:1000491]": 3920,
|
3595 |
+
"[CL:0019001]": 3921,
|
3596 |
+
"[CL:0002102]": 3922,
|
3597 |
+
"[CL:0000822]": 3923,
|
3598 |
+
"[CL:0002327]": 3924,
|
3599 |
+
"[CL:0000644]": 3925,
|
3600 |
+
"[CL:0002179]": 3926,
|
3601 |
+
"[CL:0002252]": 3927,
|
3602 |
+
"[CL:0000651]": 3928,
|
3603 |
+
"[CL:0000903]": 3929,
|
3604 |
+
"[CL:0000508]": 3930,
|
3605 |
+
"[CL:0002268]": 3931,
|
3606 |
+
"[CL:0002657]": 3932,
|
3607 |
+
"[CL:0000162]": 3933,
|
3608 |
+
"[CL:2000006]": 3934,
|
3609 |
+
"[CL:0009112]": 3935,
|
3610 |
+
"[CL:0009111]": 3936,
|
3611 |
+
"[CL:0000907]": 3937,
|
3612 |
+
"[CL:0009113]": 3938,
|
3613 |
+
"[CL:0000529]": 3939,
|
3614 |
+
"[CL:1000443]": 3940,
|
3615 |
+
"[CL:0000132]": 3941,
|
3616 |
+
"[CL:0002243]": 3942,
|
3617 |
+
"[CL:0002304]": 3943,
|
3618 |
+
"[CL:0009010]": 3944,
|
3619 |
+
"[CL:0000442]": 3945,
|
3620 |
+
"[CL:0002223]": 3946,
|
3621 |
+
"[CL:1001509]": 3947,
|
3622 |
+
"[CL:0002224]": 3948,
|
3623 |
+
"[CL:0002225]": 3949,
|
3624 |
+
"[CL:0001062]": 3950,
|
3625 |
+
"[CL:0011019]": 3951,
|
3626 |
+
"[CL:0002495]": 3952,
|
3627 |
+
"[CL:0009039]": 3953,
|
3628 |
+
"[CL:0000796]": 3954,
|
3629 |
+
"[CL:0009041]": 3955,
|
3630 |
+
"[CL:0009043]": 3956,
|
3631 |
+
"[CL:0002254]": 3957,
|
3632 |
+
"[CL:0000569]": 3958,
|
3633 |
+
"[CL:0000843]": 3959,
|
3634 |
+
"[CL:0009042]": 3960,
|
3635 |
+
"[CL:0009006]": 3961,
|
3636 |
+
"[CL:1000353]": 3962,
|
3637 |
+
"[CL:0002145]": 3963,
|
3638 |
+
"[CL:0000861]": 3964,
|
3639 |
+
"[CL:0002480]": 3965,
|
3640 |
+
"[CL:4028004]": 3966,
|
3641 |
+
"[CL:4028006]": 3967,
|
3642 |
+
"[CL:1001568]": 3968,
|
3643 |
+
"[CL:0009089]": 3969,
|
3644 |
+
"[CL:2000093]": 3970,
|
3645 |
+
"[CL:1001603]": 3971,
|
3646 |
+
"[CL:4030023]": 3972,
|
3647 |
+
"[CL:0000313]": 3973,
|
3648 |
+
"[CL:1000312]": 3974,
|
3649 |
+
"[CL:0010003]": 3975,
|
3650 |
+
"[CL:0019003]": 3976,
|
3651 |
+
"[CL:0002075]": 3977,
|
3652 |
+
"[CL:4030006]": 3978,
|
3653 |
+
"[CL:0000878]": 3979,
|
3654 |
+
"[CL:0000065]": 3980,
|
3655 |
+
"[CL:0000706]": 3981,
|
3656 |
+
"[CL:0001061]": 3982,
|
3657 |
+
"[CL:1001131]": 3983,
|
3658 |
+
"[CL:1001285]": 3984,
|
3659 |
+
"[CL:1001005]": 3985,
|
3660 |
+
"[CL:0000075]": 3986,
|
3661 |
+
"[CL:1001225]": 3987,
|
3662 |
+
"[CL:0002187]": 3988,
|
3663 |
+
"[CL:0000649]": 3989,
|
3664 |
+
"[CL:0002355]": 3990,
|
3665 |
+
"[CL:0000559]": 3991,
|
3666 |
+
"[CL:0000837]": 3992,
|
3667 |
+
"[CL:0002193]": 3993,
|
3668 |
+
"[CL:0000838]": 3994,
|
3669 |
+
"[CL:0000836]": 3995,
|
3670 |
+
"[CL:0000051]": 3996,
|
3671 |
+
"[CL:0000839]": 3997,
|
3672 |
+
"[CL:0000547]": 3998,
|
3673 |
+
"[CL:0000936]": 3999,
|
3674 |
+
"[CL:0000092]": 4000,
|
3675 |
+
"[CL:0001029]": 4001,
|
3676 |
+
"[CL:0007010]": 4002,
|
3677 |
+
"[CL:0000242]": 4003,
|
3678 |
+
"[CL:0002262]": 4004,
|
3679 |
+
"[CL:2000092]": 4005,
|
3680 |
+
"[CL:0000771]": 4006,
|
3681 |
+
"[CL:0000704]": 4007,
|
3682 |
+
"[CL:0000062]": 4008,
|
3683 |
+
"[CL:0000680]": 4009,
|
3684 |
+
"[CL:0002028]": 4010,
|
3685 |
+
"[CL:0002157]": 4011,
|
3686 |
+
"[CL:0002189]": 4012,
|
3687 |
+
"[CL:0002009]": 4013,
|
3688 |
+
"[CL:0000819]": 4014,
|
3689 |
+
"[CL:0000915]": 4015,
|
3690 |
+
"[CL:0000957]": 4016,
|
3691 |
+
"[CL:0000954]": 4017,
|
3692 |
+
"[CL:0002048]": 4018,
|
3693 |
+
"[CL:0001069]": 4019,
|
3694 |
+
"[CL:0000253]": 4020,
|
3695 |
+
"[CL:0002010]": 4021,
|
3696 |
+
"[CL:1000449]": 4022,
|
3697 |
+
"[CL:0000212]": 4023,
|
3698 |
+
"[CL:0000222]": 4024,
|
3699 |
+
"[CL:1000347]": 4025,
|
3700 |
+
"[CL:0011108]": 4026,
|
3701 |
+
"[CL:0000844]": 4027,
|
3702 |
+
"[CL:0000577]": 4028,
|
3703 |
+
"[CL:0000031]": 4029,
|
3704 |
+
"[CL:0005021]": 4030,
|
3705 |
+
"[CL:0001080]": 4031,
|
3706 |
+
"[CL:0002607]": 4032,
|
3707 |
+
"[CL:0000531]": 4033,
|
3708 |
+
"[CL:0000899]": 4034,
|
3709 |
+
"[CL:0001074]": 4035,
|
3710 |
+
"[CL:0000432]": 4036,
|
3711 |
+
"[CL:0001079]": 4037,
|
3712 |
+
"[CL:0009022]": 4038,
|
3713 |
+
"[CL:0000502]": 4039,
|
3714 |
+
"[CL:0002277]": 4040,
|
3715 |
+
"[CL:0002279]": 4041,
|
3716 |
+
"[CL:0002278]": 4042,
|
3717 |
+
"[CL:0002351]": 4043,
|
3718 |
+
"[CL:0002280]": 4044,
|
3719 |
+
"[CL:0001064]": 4045,
|
3720 |
+
"[CL:0002553]": 4046,
|
3721 |
+
"[CL:0000001]": 4047,
|
3722 |
+
"[CL:0002419]": 4048,
|
3723 |
+
"[CL:0002368]": 4049,
|
3724 |
+
"[CL:0002154]": 4050,
|
3725 |
+
"[CL:0002151]": 4051,
|
3726 |
+
"[CL:0000553]": 4052,
|
3727 |
+
"[CL:0000821]": 4053,
|
3728 |
+
"[CL:0000956]": 4054,
|
3729 |
+
"[CL:0000558]": 4055,
|
3730 |
+
"[CL:0000820]": 4056,
|
3731 |
+
"[CL:0002375]": 4057,
|
3732 |
+
"[CL:0000922]": 4058,
|
3733 |
+
"[CL:0000937]": 4059,
|
3734 |
+
"[CL:0002377]": 4060,
|
3735 |
+
"[CL:0011020]": 4061,
|
3736 |
+
"[CL:0000150]": 4062,
|
3737 |
+
"[CL:0000055]": 4063,
|
3738 |
+
"[CL:0002092]": 4064,
|
3739 |
+
"[CL:0000513]": 4065,
|
3740 |
+
"[CL:0000006]": 4066,
|
3741 |
+
"[CL:0000514]": 4067,
|
3742 |
+
"[CL:1000311]": 4068,
|
3743 |
+
"[CL:0010022]": 4069,
|
3744 |
+
"[CL:0011012]": 4070,
|
3745 |
+
"[CL:0002364]": 4071,
|
3746 |
+
"[CL:0002678]": 4072,
|
3747 |
+
"[CL:0000827]": 4073,
|
3748 |
+
"[CL:0002132]": 4074,
|
3749 |
+
"[CL:0000751]": 4075,
|
3750 |
+
"[CL:0000503]": 4076,
|
3751 |
+
"[CL:1000428]": 4077,
|
3752 |
+
"[CL:0011025]": 4078,
|
3753 |
+
"[CL:0009038]": 4079,
|
3754 |
+
"[CL:0001204]": 4080,
|
3755 |
+
"[CL:0000255]": 4081,
|
3756 |
+
"[CL:0000548]": 4082,
|
3757 |
+
"[CL:0001203]": 4083,
|
3758 |
+
"[CL:0000322]": 4084,
|
3759 |
+
"[CL:0000113]": 4085,
|
3760 |
+
"[CL:1000342]": 4086,
|
3761 |
+
"[CL:0000842]": 4087,
|
3762 |
+
"[CL:1000326]": 4088,
|
3763 |
+
"[CL:0000802]": 4089,
|
3764 |
+
"[CL:1000278]": 4090,
|
3765 |
+
"[CL:1000348]": 4091,
|
3766 |
+
"[CL:0002207]": 4092,
|
3767 |
+
"[CL:2000001]": 4093,
|
3768 |
+
"[CL:1000500]": 4094,
|
3769 |
+
"[CL:1001033]": 4095,
|
3770 |
+
"[CL:1000450]": 4096,
|
3771 |
+
"[CL:0002188]": 4097,
|
3772 |
+
"[CL:1000494]": 4098,
|
3773 |
+
"[CL:0000731]": 4099,
|
3774 |
+
"[CL:0011024]": 4100,
|
3775 |
+
"[CL:0002422]": 4101,
|
3776 |
+
"[UBERON:0002450]": 4102,
|
3777 |
+
"[UBERON:0001987]": 4103,
|
3778 |
+
"[UBERON:0000453]": 4104,
|
3779 |
+
"[UBERON:0000178]": 4105,
|
3780 |
+
"[UBERON:0002113]": 4106,
|
3781 |
+
"[UBERON:0018303]": 4107,
|
3782 |
+
"[UBERON:0005406]": 4108,
|
3783 |
+
"[UBERON:0010210]": 4109,
|
3784 |
+
"[UBERON:0000310]": 4110,
|
3785 |
+
"[UBERON:0002771]": 4111,
|
3786 |
+
"[UBERON:8410010]": 4112,
|
3787 |
+
"[UBERON:0016632]": 4113,
|
3788 |
+
"[UBERON:0001225]": 4114,
|
3789 |
+
"[UBERON:0012648]": 4115,
|
3790 |
+
"[UBERON:0001228]": 4116,
|
3791 |
+
"[UBERON:0000362]": 4117,
|
3792 |
+
"[UBERON:0003889]": 4118,
|
3793 |
+
"[UBERON:0001117]": 4119,
|
3794 |
+
"[UBERON:0008933]": 4120,
|
3795 |
+
"[UBERON:0016538]": 4121,
|
3796 |
+
"[UBERON:0002436]": 4122,
|
3797 |
+
"[UBERON:0016530]": 4123,
|
3798 |
+
"[UBERON:0001384]": 4124,
|
3799 |
+
"[UBERON:0000451]": 4125,
|
3800 |
+
"[UBERON:0002822]": 4126,
|
3801 |
+
"[UBERON:0001786]": 4127,
|
3802 |
+
"[UBERON:0013682]": 4128,
|
3803 |
+
"[UBERON:0002048]": 4129,
|
3804 |
+
"[UBERON:8410025]": 4130,
|
3805 |
+
"[UBERON:8410026]": 4131,
|
3806 |
+
"[UBERON:0009834]": 4132,
|
3807 |
+
"[UBERON:0001542]": 4133,
|
3808 |
+
"[UBERON:0003126]": 4134,
|
3809 |
+
"[UBERON:0000029]": 4135,
|
3810 |
+
"[UBERON:0002107]": 4136,
|
3811 |
+
"[UBERON:0001831]": 4137,
|
3812 |
+
"[UBERON:0002106]": 4138,
|
3813 |
+
"[UBERON:0002367]": 4139,
|
3814 |
+
"[UBERON:0002370]": 4140,
|
3815 |
+
"[UBERON:0000059]": 4141,
|
3816 |
+
"[UBERON:0001911]": 4142,
|
3817 |
+
"[UBERON:0002081]": 4143,
|
3818 |
+
"[UBERON:0010033]": 4144,
|
3819 |
+
"[UBERON:0000017]": 4145,
|
3820 |
+
"[UBERON:0001013]": 4146,
|
3821 |
+
"[UBERON:0002190]": 4147,
|
3822 |
+
"[UBERON:0001832]": 4148,
|
3823 |
+
"[UBERON:0002097]": 4149,
|
3824 |
+
"[UBERON:0002371]": 4150,
|
3825 |
+
"[UBERON:0001295]": 4151,
|
3826 |
+
"[UBERON:0000964]": 4152,
|
3827 |
+
"[UBERON:0002082]": 4153,
|
3828 |
+
"[UBERON:0018707]": 4154,
|
3829 |
+
"[UBERON:0001296]": 4155,
|
3830 |
+
"[UBERON:0000970]": 4156,
|
3831 |
+
"[UBERON:0010032]": 4157,
|
3832 |
+
"[UBERON:0001811]": 4158,
|
3833 |
+
"[UBERON:0001773]": 4159,
|
3834 |
+
"[UBERON:0008612]": 4160,
|
3835 |
+
"[UBERON:0002382]": 4161,
|
3836 |
+
"[UBERON:0003902]": 4162,
|
3837 |
+
"[UBERON:0001868]": 4163,
|
3838 |
+
"[UBERON:0001416]": 4164,
|
3839 |
+
"[UBERON:0000995]": 4165,
|
3840 |
+
"[UBERON:0001817]": 4166,
|
3841 |
+
"[UBERON:0001723]": 4167,
|
3842 |
+
"[UBERON:0002108]": 4168,
|
3843 |
+
"[UBERON:0002378]": 4169,
|
3844 |
+
"[UBERON:0001621]": 4170,
|
3845 |
+
"[UBERON:0002385]": 4171,
|
3846 |
+
"[UBERON:0002049]": 4172,
|
3847 |
+
"[UBERON:0000947]": 4173,
|
3848 |
+
"[UBERON:0000016]": 4174,
|
3849 |
+
"[UBERON:0008946]": 4175,
|
3850 |
+
"[UBERON:0001005]": 4176,
|
3851 |
+
"[UBERON:0000004]": 4177,
|
3852 |
+
"[UBERON:0002037]": 4178,
|
3853 |
+
"[UBERON:0001836]": 4179,
|
3854 |
+
"[UBERON:0000991]": 4180,
|
3855 |
+
"[UBERON:0001893]": 4181,
|
3856 |
+
"[UBERON:0000948]": 4182,
|
3857 |
+
"[UBERON:0000160]": 4183,
|
3858 |
+
"[UBERON:0001264]": 4184,
|
3859 |
+
"[UBERON:0001630]": 4185,
|
3860 |
+
"[UBERON:0002369]": 4186,
|
3861 |
+
"[UBERON:0000945]": 4187,
|
3862 |
+
"[UBERON:0035328]": 4188,
|
3863 |
+
"[UBERON:0005616]": 4189,
|
3864 |
+
"[UBERON:0001155]": 4190,
|
3865 |
+
"[UBERON:0002116]": 4191,
|
3866 |
+
"[UBERON:0002098]": 4192,
|
3867 |
+
"[UBERON:0002079]": 4193,
|
3868 |
+
"[UBERON:0002084]": 4194,
|
3869 |
+
"[UBERON:0002080]": 4195,
|
3870 |
+
"[UBERON:0002094]": 4196,
|
3871 |
+
"[UBERON:0001046]": 4197,
|
3872 |
+
"[UBERON:0002078]": 4198,
|
3873 |
+
"[UBERON:0002114]": 4199,
|
3874 |
+
"[UBERON:0001043]": 4200,
|
3875 |
+
"[UBERON:0001161]": 4201,
|
3876 |
+
"[UBERON:0002115]": 4202,
|
3877 |
+
"[UBERON:0001165]": 4203,
|
3878 |
+
"[UBERON:0001901]": 4204,
|
3879 |
+
"[UBERON:0002429]": 4205,
|
3880 |
+
"[UBERON:0000473]": 4206,
|
3881 |
+
"[UBERON:0000955]": 4207,
|
3882 |
+
"[UBERON:0007106]": 4208,
|
3883 |
+
"[UBERON:0012168]": 4209,
|
3884 |
+
"[UBERON:0004339]": 4210,
|
3885 |
+
"[UBERON:0000992]": 4211,
|
3886 |
+
"[UBERON:0000056]": 4212,
|
3887 |
+
"[UBERON:0000006]": 4213,
|
3888 |
+
"[UBERON:0000956]": 4214,
|
3889 |
+
"[UBERON:0002661]": 4215,
|
3890 |
+
"[UBERON:0013756]": 4216,
|
3891 |
+
"[UBERON:0039167]": 4217,
|
3892 |
+
"[UBERON:0008953]": 4218,
|
3893 |
+
"[UBERON:0002728]": 4219,
|
3894 |
+
"[UBERON:0001637]": 4220,
|
3895 |
+
"[UBERON:0000002]": 4221,
|
3896 |
+
"[UBERON:0001157]": 4222,
|
3897 |
+
"[UBERON:0001154]": 4223,
|
3898 |
+
"[UBERON:0001156]": 4224,
|
3899 |
+
"[UBERON:0000977]": 4225,
|
3900 |
+
"[UBERON:0002046]": 4226,
|
3901 |
+
"[UBERON:0003688]": 4227,
|
3902 |
+
"[UBERON:0001871]": 4228,
|
3903 |
+
"[UBERON:0001052]": 4229,
|
3904 |
+
"[UBERON:0001159]": 4230,
|
3905 |
+
"[UBERON:0002110]": 4231,
|
3906 |
+
"[UBERON:0002228]": 4232,
|
3907 |
+
"[UBERON:0002240]": 4233,
|
3908 |
+
"[UBERON:0001898]": 4234,
|
3909 |
+
"[UBERON:8440012]": 4235,
|
3910 |
+
"[UBERON:0010225]": 4236,
|
3911 |
+
"[UBERON:0000988]": 4237,
|
3912 |
+
"[UBERON:0002421]": 4238,
|
3913 |
+
"[UBERON:0001891]": 4239,
|
3914 |
+
"[UBERON:0005290]": 4240,
|
3915 |
+
"[UBERON:0008972]": 4241,
|
3916 |
+
"[UBERON:0007177]": 4242,
|
3917 |
+
"[UBERON:0001153]": 4243,
|
3918 |
+
"[UBERON:0008971]": 4244,
|
3919 |
+
"[UBERON:0000397]": 4245,
|
3920 |
+
"[UBERON:0005636]": 4246,
|
3921 |
+
"[UBERON:0000966]": 4247,
|
3922 |
+
"[UBERON:0001238]": 4248,
|
3923 |
+
"[UBERON:0008345]": 4249,
|
3924 |
+
"[UBERON:0013473]": 4250,
|
3925 |
+
"[UBERON:0001162]": 4251,
|
3926 |
+
"[UBERON:0007650]": 4252,
|
3927 |
+
"[UBERON:0008989]": 4253,
|
3928 |
+
"[UBERON:0002372]": 4254,
|
3929 |
+
"[UBERON:0001769]": 4255,
|
3930 |
+
"[UBERON:0006761]": 4256,
|
3931 |
+
"[UBERON:0005969]": 4257,
|
3932 |
+
"[UBERON:0001775]": 4258,
|
3933 |
+
"[UBERON:0000965]": 4259,
|
3934 |
+
"[UBERON:0002299]": 4260,
|
3935 |
+
"[UBERON:0035213]": 4261,
|
3936 |
+
"[UBERON:0001158]": 4262,
|
3937 |
+
"[UBERON:0002358]": 4263,
|
3938 |
+
"[UBERON:0036292]": 4264,
|
3939 |
+
"[UBERON:0007795]": 4265,
|
3940 |
+
"[UBERON:0002119]": 4266,
|
3941 |
+
"[UBERON:0002118]": 4267,
|
3942 |
+
"[UBERON:0000916]": 4268,
|
3943 |
+
"[UBERON:0001103]": 4269,
|
3944 |
+
"[UBERON:0003697]": 4270,
|
3945 |
+
"[UBERON:0035210]": 4271,
|
3946 |
+
"[UBERON:0001366]": 4272,
|
3947 |
+
"[UBERON:0001255]": 4273,
|
3948 |
+
"[UBERON:0008952]": 4274,
|
3949 |
+
"[UBERON:0002174]": 4275,
|
3950 |
+
"[UBERON:0001332]": 4276,
|
3951 |
+
"[UBERON:0002100]": 4277,
|
3952 |
+
"[UBERON:0000403]": 4278,
|
3953 |
+
"[UBERON:0000328]": 4279,
|
3954 |
+
"[UBERON:0000053]": 4280,
|
3955 |
+
"[UBERON:0001040]": 4281,
|
3956 |
+
"[UBERON:0002509]": 4282,
|
3957 |
+
"[UBERON:0012474]": 4283,
|
3958 |
+
"[UBERON:0022277]": 4284,
|
3959 |
+
"[UBERON:0000175]": 4285,
|
3960 |
+
"[UBERON:0001890]": 4286,
|
3961 |
+
"[UBERON:0002811]": 4287,
|
3962 |
+
"[UBERON:0002808]": 4288,
|
3963 |
+
"[UBERON:0002810]": 4289,
|
3964 |
+
"[UBERON:0002803]": 4290,
|
3965 |
+
"[UBERON:0002809]": 4291,
|
3966 |
+
"[UBERON:0023852]": 4292,
|
3967 |
+
"[UBERON:0002802]": 4293,
|
3968 |
+
"[UBERON:0016525]": 4294,
|
3969 |
+
"[UBERON:0002021]": 4295,
|
3970 |
+
"[UBERON:0001872]": 4296,
|
3971 |
+
"[UBERON:8410000]": 4297,
|
3972 |
+
"[UBERON:0007625]": 4298,
|
3973 |
+
"[UBERON:0000014]": 4299,
|
3974 |
+
"[UBERON:0009472]": 4300,
|
3975 |
+
"[UBERON:0013706]": 4301,
|
3976 |
+
"[UBERON:0007644]": 4302,
|
3977 |
+
"[UBERON:0000030]": 4303,
|
3978 |
+
"[UBERON:0000400]": 4304,
|
3979 |
+
"[UBERON:0001134]": 4305,
|
3980 |
+
"[UBERON:0001976]": 4306,
|
3981 |
+
"[UBERON:0001707]": 4307,
|
3982 |
+
"[UBERON:0002185]": 4308,
|
3983 |
+
"[UBERON:0001224]": 4309,
|
3984 |
+
"[UBERON:0003517]": 4310
|
3985 |
+
},
|
3986 |
+
"unk_token": "<UNK>"
|
3987 |
+
}
|
3988 |
+
}
|
tokenizer/config.yaml
ADDED
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
tokenizers_info:
|
2 |
+
- name: AA
|
3 |
+
tokenizer_id: 0
|
4 |
+
json_path: ./t5_tokenizer_AA_special.json
|
5 |
+
modular_json_path: ./t5_tokenizer_AA_special.json
|
6 |
+
start_delimiter: <start_AA>
|
7 |
+
end_delimiter: <end_AA>
|
8 |
+
- name: SMILES
|
9 |
+
tokenizer_id: 1
|
10 |
+
json_path: ./bpe_tokenizer_trained_on_chembl_zinc_with_aug_4272372_samples_balanced_1_1.json
|
11 |
+
modular_json_path: ./bpe_tokenizer_trained_on_chembl_zinc_with_aug_4272372_samples_balanced_1_1.json
|
12 |
+
start_delimiter: <start_SMILES>
|
13 |
+
end_delimiter: <end_SMILES>
|
14 |
+
- name: CELL_ATTRIBUTES
|
15 |
+
tokenizer_id: 2
|
16 |
+
json_path: ./cell_attributes_tokenizer.json
|
17 |
+
modular_json_path: ./cell_attributes_tokenizer.json
|
18 |
+
start_delimiter: <start_CELL_ATTRIBUTES>
|
19 |
+
end_delimiter: <end_CELL_ATTRIBUTES>
|
20 |
+
- name: GENE
|
21 |
+
tokenizer_id: 3
|
22 |
+
json_path: ./gene_tokenizer.json
|
23 |
+
modular_json_path: ./gene_tokenizer.json
|
24 |
+
start_delimiter: <start_GENE>
|
25 |
+
end_delimiter: <end_GENE>
|
26 |
+
minimal_token_id: 5000
|
27 |
+
max_possible_token_id: 100000
|
28 |
+
max_special_token_id: 500
|
tokenizer/gene_tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer/t5_tokenizer_AA_special.json
ADDED
@@ -0,0 +1,3225 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"version": "1.0",
|
3 |
+
"truncation": null,
|
4 |
+
"padding": null,
|
5 |
+
"added_tokens": [
|
6 |
+
{
|
7 |
+
"id": 0,
|
8 |
+
"content": "<UNK>",
|
9 |
+
"single_word": false,
|
10 |
+
"lstrip": false,
|
11 |
+
"rstrip": false,
|
12 |
+
"normalized": false,
|
13 |
+
"special": true
|
14 |
+
},
|
15 |
+
{
|
16 |
+
"id": 1,
|
17 |
+
"content": "<PAD>",
|
18 |
+
"single_word": false,
|
19 |
+
"lstrip": false,
|
20 |
+
"rstrip": false,
|
21 |
+
"normalized": false,
|
22 |
+
"special": true
|
23 |
+
},
|
24 |
+
{
|
25 |
+
"id": 2,
|
26 |
+
"content": "<CLS>",
|
27 |
+
"single_word": false,
|
28 |
+
"lstrip": false,
|
29 |
+
"rstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"special": true
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"id": 3,
|
35 |
+
"content": "<SEP>",
|
36 |
+
"single_word": false,
|
37 |
+
"lstrip": false,
|
38 |
+
"rstrip": false,
|
39 |
+
"normalized": false,
|
40 |
+
"special": true
|
41 |
+
},
|
42 |
+
{
|
43 |
+
"id": 4,
|
44 |
+
"content": "<MASK>",
|
45 |
+
"single_word": false,
|
46 |
+
"lstrip": false,
|
47 |
+
"rstrip": false,
|
48 |
+
"normalized": false,
|
49 |
+
"special": true
|
50 |
+
},
|
51 |
+
{
|
52 |
+
"id": 5,
|
53 |
+
"content": "<EOS>",
|
54 |
+
"single_word": false,
|
55 |
+
"lstrip": false,
|
56 |
+
"rstrip": false,
|
57 |
+
"normalized": false,
|
58 |
+
"special": true
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"id": 6,
|
62 |
+
"content": "<MOLECULAR_ENTITY>",
|
63 |
+
"single_word": false,
|
64 |
+
"lstrip": false,
|
65 |
+
"rstrip": false,
|
66 |
+
"normalized": false,
|
67 |
+
"special": true
|
68 |
+
},
|
69 |
+
{
|
70 |
+
"id": 7,
|
71 |
+
"content": "<GLOBAL_INTERACTION_ATTRIBUTES>",
|
72 |
+
"single_word": false,
|
73 |
+
"lstrip": false,
|
74 |
+
"rstrip": false,
|
75 |
+
"normalized": false,
|
76 |
+
"special": true
|
77 |
+
},
|
78 |
+
{
|
79 |
+
"id": 8,
|
80 |
+
"content": "<MOLECULAR_ENTITY_ANTIGEN>",
|
81 |
+
"single_word": false,
|
82 |
+
"lstrip": false,
|
83 |
+
"rstrip": false,
|
84 |
+
"normalized": false,
|
85 |
+
"special": true
|
86 |
+
},
|
87 |
+
{
|
88 |
+
"id": 9,
|
89 |
+
"content": "<MOLECULAR_ENTITY_EPITOPE>",
|
90 |
+
"single_word": false,
|
91 |
+
"lstrip": false,
|
92 |
+
"rstrip": false,
|
93 |
+
"normalized": false,
|
94 |
+
"special": true
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"id": 10,
|
98 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN>",
|
99 |
+
"single_word": false,
|
100 |
+
"lstrip": false,
|
101 |
+
"rstrip": false,
|
102 |
+
"normalized": false,
|
103 |
+
"special": true
|
104 |
+
},
|
105 |
+
{
|
106 |
+
"id": 11,
|
107 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN>",
|
108 |
+
"single_word": false,
|
109 |
+
"lstrip": false,
|
110 |
+
"rstrip": false,
|
111 |
+
"normalized": false,
|
112 |
+
"special": true
|
113 |
+
},
|
114 |
+
{
|
115 |
+
"id": 12,
|
116 |
+
"content": "<MOLECULAR_ENTITY_TCR_ALPHA_CHAIN>",
|
117 |
+
"single_word": false,
|
118 |
+
"lstrip": false,
|
119 |
+
"rstrip": false,
|
120 |
+
"normalized": false,
|
121 |
+
"special": true
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"id": 13,
|
125 |
+
"content": "<MOLECULAR_ENTITY_TCR_BETA_VDJ>",
|
126 |
+
"single_word": false,
|
127 |
+
"lstrip": false,
|
128 |
+
"rstrip": false,
|
129 |
+
"normalized": false,
|
130 |
+
"special": true
|
131 |
+
},
|
132 |
+
{
|
133 |
+
"id": 14,
|
134 |
+
"content": "<MOLECULAR_ENTITY_TCR_BETA_CDR3>",
|
135 |
+
"single_word": false,
|
136 |
+
"lstrip": false,
|
137 |
+
"rstrip": false,
|
138 |
+
"normalized": false,
|
139 |
+
"special": true
|
140 |
+
},
|
141 |
+
{
|
142 |
+
"id": 15,
|
143 |
+
"content": "<BINDING_AFFINITY_CLASS>",
|
144 |
+
"single_word": false,
|
145 |
+
"lstrip": false,
|
146 |
+
"rstrip": false,
|
147 |
+
"normalized": false,
|
148 |
+
"special": true
|
149 |
+
},
|
150 |
+
{
|
151 |
+
"id": 16,
|
152 |
+
"content": "<DECODER_START>",
|
153 |
+
"single_word": false,
|
154 |
+
"lstrip": false,
|
155 |
+
"rstrip": false,
|
156 |
+
"normalized": false,
|
157 |
+
"special": true
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"id": 17,
|
161 |
+
"content": "<BINDING>",
|
162 |
+
"single_word": false,
|
163 |
+
"lstrip": false,
|
164 |
+
"rstrip": false,
|
165 |
+
"normalized": false,
|
166 |
+
"special": true
|
167 |
+
},
|
168 |
+
{
|
169 |
+
"id": 18,
|
170 |
+
"content": "<FILLIN>",
|
171 |
+
"single_word": false,
|
172 |
+
"lstrip": false,
|
173 |
+
"rstrip": false,
|
174 |
+
"normalized": false,
|
175 |
+
"special": true
|
176 |
+
},
|
177 |
+
{
|
178 |
+
"id": 19,
|
179 |
+
"content": "<REORDER>",
|
180 |
+
"single_word": false,
|
181 |
+
"lstrip": false,
|
182 |
+
"rstrip": false,
|
183 |
+
"normalized": false,
|
184 |
+
"special": true
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"id": 20,
|
188 |
+
"content": "<TOAA>",
|
189 |
+
"single_word": false,
|
190 |
+
"lstrip": false,
|
191 |
+
"rstrip": false,
|
192 |
+
"normalized": false,
|
193 |
+
"special": true
|
194 |
+
},
|
195 |
+
{
|
196 |
+
"id": 21,
|
197 |
+
"content": "<ACTIVE>",
|
198 |
+
"single_word": false,
|
199 |
+
"lstrip": false,
|
200 |
+
"rstrip": false,
|
201 |
+
"normalized": false,
|
202 |
+
"special": true
|
203 |
+
},
|
204 |
+
{
|
205 |
+
"id": 22,
|
206 |
+
"content": "<GENESEQ>",
|
207 |
+
"single_word": false,
|
208 |
+
"lstrip": false,
|
209 |
+
"rstrip": false,
|
210 |
+
"normalized": false,
|
211 |
+
"special": true
|
212 |
+
},
|
213 |
+
{
|
214 |
+
"id": 23,
|
215 |
+
"content": "<INCREASE>",
|
216 |
+
"single_word": false,
|
217 |
+
"lstrip": false,
|
218 |
+
"rstrip": false,
|
219 |
+
"normalized": false,
|
220 |
+
"special": true
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"id": 24,
|
224 |
+
"content": "<DECREASE>",
|
225 |
+
"single_word": false,
|
226 |
+
"lstrip": false,
|
227 |
+
"rstrip": false,
|
228 |
+
"normalized": false,
|
229 |
+
"special": true
|
230 |
+
},
|
231 |
+
{
|
232 |
+
"id": 25,
|
233 |
+
"content": "<STRUCTURE>",
|
234 |
+
"single_word": false,
|
235 |
+
"lstrip": false,
|
236 |
+
"rstrip": false,
|
237 |
+
"normalized": false,
|
238 |
+
"special": true
|
239 |
+
},
|
240 |
+
{
|
241 |
+
"id": 26,
|
242 |
+
"content": "<DISTANCE>",
|
243 |
+
"single_word": false,
|
244 |
+
"lstrip": false,
|
245 |
+
"rstrip": false,
|
246 |
+
"normalized": false,
|
247 |
+
"special": true
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"id": 27,
|
251 |
+
"content": "<SOLUBILITY>",
|
252 |
+
"single_word": false,
|
253 |
+
"lstrip": false,
|
254 |
+
"rstrip": false,
|
255 |
+
"normalized": false,
|
256 |
+
"special": true
|
257 |
+
},
|
258 |
+
{
|
259 |
+
"id": 28,
|
260 |
+
"content": "<TOXICITY>",
|
261 |
+
"single_word": false,
|
262 |
+
"lstrip": false,
|
263 |
+
"rstrip": false,
|
264 |
+
"normalized": false,
|
265 |
+
"special": true
|
266 |
+
},
|
267 |
+
{
|
268 |
+
"id": 29,
|
269 |
+
"content": "<AB>",
|
270 |
+
"single_word": false,
|
271 |
+
"lstrip": false,
|
272 |
+
"rstrip": false,
|
273 |
+
"normalized": false,
|
274 |
+
"special": true
|
275 |
+
},
|
276 |
+
{
|
277 |
+
"id": 30,
|
278 |
+
"content": "<ISACTIVE>",
|
279 |
+
"single_word": false,
|
280 |
+
"lstrip": false,
|
281 |
+
"rstrip": false,
|
282 |
+
"normalized": false,
|
283 |
+
"special": true
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"id": 31,
|
287 |
+
"content": "<ISSYNTHETIC>",
|
288 |
+
"single_word": false,
|
289 |
+
"lstrip": false,
|
290 |
+
"rstrip": false,
|
291 |
+
"normalized": false,
|
292 |
+
"special": true
|
293 |
+
},
|
294 |
+
{
|
295 |
+
"id": 32,
|
296 |
+
"content": "<PENETR>",
|
297 |
+
"single_word": false,
|
298 |
+
"lstrip": false,
|
299 |
+
"rstrip": false,
|
300 |
+
"normalized": false,
|
301 |
+
"special": true
|
302 |
+
},
|
303 |
+
{
|
304 |
+
"id": 33,
|
305 |
+
"content": "<ABSORPTION>",
|
306 |
+
"single_word": false,
|
307 |
+
"lstrip": false,
|
308 |
+
"rstrip": false,
|
309 |
+
"normalized": false,
|
310 |
+
"special": true
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"id": 34,
|
314 |
+
"content": "<DISTRIBUTION>",
|
315 |
+
"single_word": false,
|
316 |
+
"lstrip": false,
|
317 |
+
"rstrip": false,
|
318 |
+
"normalized": false,
|
319 |
+
"special": true
|
320 |
+
},
|
321 |
+
{
|
322 |
+
"id": 35,
|
323 |
+
"content": "<METABOLISM>",
|
324 |
+
"single_word": false,
|
325 |
+
"lstrip": false,
|
326 |
+
"rstrip": false,
|
327 |
+
"normalized": false,
|
328 |
+
"special": true
|
329 |
+
},
|
330 |
+
{
|
331 |
+
"id": 36,
|
332 |
+
"content": "<EXCRETION>",
|
333 |
+
"single_word": false,
|
334 |
+
"lstrip": false,
|
335 |
+
"rstrip": false,
|
336 |
+
"normalized": false,
|
337 |
+
"special": true
|
338 |
+
},
|
339 |
+
{
|
340 |
+
"id": 37,
|
341 |
+
"content": "<FLUORESCENCE>",
|
342 |
+
"single_word": false,
|
343 |
+
"lstrip": false,
|
344 |
+
"rstrip": false,
|
345 |
+
"normalized": false,
|
346 |
+
"special": true
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"id": 38,
|
350 |
+
"content": "<STABILITY>",
|
351 |
+
"single_word": false,
|
352 |
+
"lstrip": false,
|
353 |
+
"rstrip": false,
|
354 |
+
"normalized": false,
|
355 |
+
"special": true
|
356 |
+
},
|
357 |
+
{
|
358 |
+
"id": 39,
|
359 |
+
"content": "<DISORDER>",
|
360 |
+
"single_word": false,
|
361 |
+
"lstrip": false,
|
362 |
+
"rstrip": false,
|
363 |
+
"normalized": false,
|
364 |
+
"special": true
|
365 |
+
},
|
366 |
+
{
|
367 |
+
"id": 40,
|
368 |
+
"content": "<DISEASE>",
|
369 |
+
"single_word": false,
|
370 |
+
"lstrip": false,
|
371 |
+
"rstrip": false,
|
372 |
+
"normalized": false,
|
373 |
+
"special": true
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"id": 41,
|
377 |
+
"content": "<BINARY>",
|
378 |
+
"single_word": false,
|
379 |
+
"lstrip": false,
|
380 |
+
"rstrip": false,
|
381 |
+
"normalized": false,
|
382 |
+
"special": true
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"id": 42,
|
386 |
+
"content": "<REGRESSION>",
|
387 |
+
"single_word": false,
|
388 |
+
"lstrip": false,
|
389 |
+
"rstrip": false,
|
390 |
+
"normalized": false,
|
391 |
+
"special": true
|
392 |
+
},
|
393 |
+
{
|
394 |
+
"id": 43,
|
395 |
+
"content": "<ORGANISM>",
|
396 |
+
"single_word": false,
|
397 |
+
"lstrip": false,
|
398 |
+
"rstrip": false,
|
399 |
+
"normalized": false,
|
400 |
+
"special": true
|
401 |
+
},
|
402 |
+
{
|
403 |
+
"id": 44,
|
404 |
+
"content": "<0>",
|
405 |
+
"single_word": false,
|
406 |
+
"lstrip": false,
|
407 |
+
"rstrip": false,
|
408 |
+
"normalized": false,
|
409 |
+
"special": true
|
410 |
+
},
|
411 |
+
{
|
412 |
+
"id": 45,
|
413 |
+
"content": "<1>",
|
414 |
+
"single_word": false,
|
415 |
+
"lstrip": false,
|
416 |
+
"rstrip": false,
|
417 |
+
"normalized": false,
|
418 |
+
"special": true
|
419 |
+
},
|
420 |
+
{
|
421 |
+
"id": 46,
|
422 |
+
"content": "<2>",
|
423 |
+
"single_word": false,
|
424 |
+
"lstrip": false,
|
425 |
+
"rstrip": false,
|
426 |
+
"normalized": false,
|
427 |
+
"special": true
|
428 |
+
},
|
429 |
+
{
|
430 |
+
"id": 47,
|
431 |
+
"content": "<3>",
|
432 |
+
"single_word": false,
|
433 |
+
"lstrip": false,
|
434 |
+
"rstrip": false,
|
435 |
+
"normalized": false,
|
436 |
+
"special": true
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"id": 48,
|
440 |
+
"content": "<4>",
|
441 |
+
"single_word": false,
|
442 |
+
"lstrip": false,
|
443 |
+
"rstrip": false,
|
444 |
+
"normalized": false,
|
445 |
+
"special": true
|
446 |
+
},
|
447 |
+
{
|
448 |
+
"id": 49,
|
449 |
+
"content": "<5>",
|
450 |
+
"single_word": false,
|
451 |
+
"lstrip": false,
|
452 |
+
"rstrip": false,
|
453 |
+
"normalized": false,
|
454 |
+
"special": true
|
455 |
+
},
|
456 |
+
{
|
457 |
+
"id": 50,
|
458 |
+
"content": "<6>",
|
459 |
+
"single_word": false,
|
460 |
+
"lstrip": false,
|
461 |
+
"rstrip": false,
|
462 |
+
"normalized": false,
|
463 |
+
"special": true
|
464 |
+
},
|
465 |
+
{
|
466 |
+
"id": 51,
|
467 |
+
"content": "<7>",
|
468 |
+
"single_word": false,
|
469 |
+
"lstrip": false,
|
470 |
+
"rstrip": false,
|
471 |
+
"normalized": false,
|
472 |
+
"special": true
|
473 |
+
},
|
474 |
+
{
|
475 |
+
"id": 52,
|
476 |
+
"content": "<8>",
|
477 |
+
"single_word": false,
|
478 |
+
"lstrip": false,
|
479 |
+
"rstrip": false,
|
480 |
+
"normalized": false,
|
481 |
+
"special": true
|
482 |
+
},
|
483 |
+
{
|
484 |
+
"id": 53,
|
485 |
+
"content": "<9>",
|
486 |
+
"single_word": false,
|
487 |
+
"lstrip": false,
|
488 |
+
"rstrip": false,
|
489 |
+
"normalized": false,
|
490 |
+
"special": true
|
491 |
+
},
|
492 |
+
{
|
493 |
+
"id": 54,
|
494 |
+
"content": "<.>",
|
495 |
+
"single_word": false,
|
496 |
+
"lstrip": false,
|
497 |
+
"rstrip": false,
|
498 |
+
"normalized": false,
|
499 |
+
"special": true
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"id": 55,
|
503 |
+
"content": "<YES>",
|
504 |
+
"single_word": false,
|
505 |
+
"lstrip": false,
|
506 |
+
"rstrip": false,
|
507 |
+
"normalized": false,
|
508 |
+
"special": true
|
509 |
+
},
|
510 |
+
{
|
511 |
+
"id": 56,
|
512 |
+
"content": "<NO>",
|
513 |
+
"single_word": false,
|
514 |
+
"lstrip": false,
|
515 |
+
"rstrip": false,
|
516 |
+
"normalized": false,
|
517 |
+
"special": true
|
518 |
+
},
|
519 |
+
{
|
520 |
+
"id": 57,
|
521 |
+
"content": "<SENTINEL_ID_0>",
|
522 |
+
"single_word": false,
|
523 |
+
"lstrip": false,
|
524 |
+
"rstrip": false,
|
525 |
+
"normalized": false,
|
526 |
+
"special": true
|
527 |
+
},
|
528 |
+
{
|
529 |
+
"id": 58,
|
530 |
+
"content": "<SENTINEL_ID_1>",
|
531 |
+
"single_word": false,
|
532 |
+
"lstrip": false,
|
533 |
+
"rstrip": false,
|
534 |
+
"normalized": false,
|
535 |
+
"special": true
|
536 |
+
},
|
537 |
+
{
|
538 |
+
"id": 59,
|
539 |
+
"content": "<SENTINEL_ID_2>",
|
540 |
+
"single_word": false,
|
541 |
+
"lstrip": false,
|
542 |
+
"rstrip": false,
|
543 |
+
"normalized": false,
|
544 |
+
"special": true
|
545 |
+
},
|
546 |
+
{
|
547 |
+
"id": 60,
|
548 |
+
"content": "<SENTINEL_ID_3>",
|
549 |
+
"single_word": false,
|
550 |
+
"lstrip": false,
|
551 |
+
"rstrip": false,
|
552 |
+
"normalized": false,
|
553 |
+
"special": true
|
554 |
+
},
|
555 |
+
{
|
556 |
+
"id": 61,
|
557 |
+
"content": "<SENTINEL_ID_4>",
|
558 |
+
"single_word": false,
|
559 |
+
"lstrip": false,
|
560 |
+
"rstrip": false,
|
561 |
+
"normalized": false,
|
562 |
+
"special": true
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"id": 62,
|
566 |
+
"content": "<SENTINEL_ID_5>",
|
567 |
+
"single_word": false,
|
568 |
+
"lstrip": false,
|
569 |
+
"rstrip": false,
|
570 |
+
"normalized": false,
|
571 |
+
"special": true
|
572 |
+
},
|
573 |
+
{
|
574 |
+
"id": 63,
|
575 |
+
"content": "<SENTINEL_ID_6>",
|
576 |
+
"single_word": false,
|
577 |
+
"lstrip": false,
|
578 |
+
"rstrip": false,
|
579 |
+
"normalized": false,
|
580 |
+
"special": true
|
581 |
+
},
|
582 |
+
{
|
583 |
+
"id": 64,
|
584 |
+
"content": "<SENTINEL_ID_7>",
|
585 |
+
"single_word": false,
|
586 |
+
"lstrip": false,
|
587 |
+
"rstrip": false,
|
588 |
+
"normalized": false,
|
589 |
+
"special": true
|
590 |
+
},
|
591 |
+
{
|
592 |
+
"id": 65,
|
593 |
+
"content": "<SENTINEL_ID_8>",
|
594 |
+
"single_word": false,
|
595 |
+
"lstrip": false,
|
596 |
+
"rstrip": false,
|
597 |
+
"normalized": false,
|
598 |
+
"special": true
|
599 |
+
},
|
600 |
+
{
|
601 |
+
"id": 66,
|
602 |
+
"content": "<SENTINEL_ID_9>",
|
603 |
+
"single_word": false,
|
604 |
+
"lstrip": false,
|
605 |
+
"rstrip": false,
|
606 |
+
"normalized": false,
|
607 |
+
"special": true
|
608 |
+
},
|
609 |
+
{
|
610 |
+
"id": 67,
|
611 |
+
"content": "<SENTINEL_ID_10>",
|
612 |
+
"single_word": false,
|
613 |
+
"lstrip": false,
|
614 |
+
"rstrip": false,
|
615 |
+
"normalized": false,
|
616 |
+
"special": true
|
617 |
+
},
|
618 |
+
{
|
619 |
+
"id": 68,
|
620 |
+
"content": "<SENTINEL_ID_11>",
|
621 |
+
"single_word": false,
|
622 |
+
"lstrip": false,
|
623 |
+
"rstrip": false,
|
624 |
+
"normalized": false,
|
625 |
+
"special": true
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"id": 69,
|
629 |
+
"content": "<SENTINEL_ID_12>",
|
630 |
+
"single_word": false,
|
631 |
+
"lstrip": false,
|
632 |
+
"rstrip": false,
|
633 |
+
"normalized": false,
|
634 |
+
"special": true
|
635 |
+
},
|
636 |
+
{
|
637 |
+
"id": 70,
|
638 |
+
"content": "<SENTINEL_ID_13>",
|
639 |
+
"single_word": false,
|
640 |
+
"lstrip": false,
|
641 |
+
"rstrip": false,
|
642 |
+
"normalized": false,
|
643 |
+
"special": true
|
644 |
+
},
|
645 |
+
{
|
646 |
+
"id": 71,
|
647 |
+
"content": "<SENTINEL_ID_14>",
|
648 |
+
"single_word": false,
|
649 |
+
"lstrip": false,
|
650 |
+
"rstrip": false,
|
651 |
+
"normalized": false,
|
652 |
+
"special": true
|
653 |
+
},
|
654 |
+
{
|
655 |
+
"id": 72,
|
656 |
+
"content": "<SENTINEL_ID_15>",
|
657 |
+
"single_word": false,
|
658 |
+
"lstrip": false,
|
659 |
+
"rstrip": false,
|
660 |
+
"normalized": false,
|
661 |
+
"special": true
|
662 |
+
},
|
663 |
+
{
|
664 |
+
"id": 73,
|
665 |
+
"content": "<SENTINEL_ID_16>",
|
666 |
+
"single_word": false,
|
667 |
+
"lstrip": false,
|
668 |
+
"rstrip": false,
|
669 |
+
"normalized": false,
|
670 |
+
"special": true
|
671 |
+
},
|
672 |
+
{
|
673 |
+
"id": 74,
|
674 |
+
"content": "<SENTINEL_ID_17>",
|
675 |
+
"single_word": false,
|
676 |
+
"lstrip": false,
|
677 |
+
"rstrip": false,
|
678 |
+
"normalized": false,
|
679 |
+
"special": true
|
680 |
+
},
|
681 |
+
{
|
682 |
+
"id": 75,
|
683 |
+
"content": "<SENTINEL_ID_18>",
|
684 |
+
"single_word": false,
|
685 |
+
"lstrip": false,
|
686 |
+
"rstrip": false,
|
687 |
+
"normalized": false,
|
688 |
+
"special": true
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"id": 76,
|
692 |
+
"content": "<SENTINEL_ID_19>",
|
693 |
+
"single_word": false,
|
694 |
+
"lstrip": false,
|
695 |
+
"rstrip": false,
|
696 |
+
"normalized": false,
|
697 |
+
"special": true
|
698 |
+
},
|
699 |
+
{
|
700 |
+
"id": 77,
|
701 |
+
"content": "<SENTINEL_ID_20>",
|
702 |
+
"single_word": false,
|
703 |
+
"lstrip": false,
|
704 |
+
"rstrip": false,
|
705 |
+
"normalized": false,
|
706 |
+
"special": true
|
707 |
+
},
|
708 |
+
{
|
709 |
+
"id": 78,
|
710 |
+
"content": "<SENTINEL_ID_21>",
|
711 |
+
"single_word": false,
|
712 |
+
"lstrip": false,
|
713 |
+
"rstrip": false,
|
714 |
+
"normalized": false,
|
715 |
+
"special": true
|
716 |
+
},
|
717 |
+
{
|
718 |
+
"id": 79,
|
719 |
+
"content": "<SENTINEL_ID_22>",
|
720 |
+
"single_word": false,
|
721 |
+
"lstrip": false,
|
722 |
+
"rstrip": false,
|
723 |
+
"normalized": false,
|
724 |
+
"special": true
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"id": 80,
|
728 |
+
"content": "<SENTINEL_ID_23>",
|
729 |
+
"single_word": false,
|
730 |
+
"lstrip": false,
|
731 |
+
"rstrip": false,
|
732 |
+
"normalized": false,
|
733 |
+
"special": true
|
734 |
+
},
|
735 |
+
{
|
736 |
+
"id": 81,
|
737 |
+
"content": "<SENTINEL_ID_24>",
|
738 |
+
"single_word": false,
|
739 |
+
"lstrip": false,
|
740 |
+
"rstrip": false,
|
741 |
+
"normalized": false,
|
742 |
+
"special": true
|
743 |
+
},
|
744 |
+
{
|
745 |
+
"id": 82,
|
746 |
+
"content": "<SENTINEL_ID_25>",
|
747 |
+
"single_word": false,
|
748 |
+
"lstrip": false,
|
749 |
+
"rstrip": false,
|
750 |
+
"normalized": false,
|
751 |
+
"special": true
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"id": 83,
|
755 |
+
"content": "<SENTINEL_ID_26>",
|
756 |
+
"single_word": false,
|
757 |
+
"lstrip": false,
|
758 |
+
"rstrip": false,
|
759 |
+
"normalized": false,
|
760 |
+
"special": true
|
761 |
+
},
|
762 |
+
{
|
763 |
+
"id": 84,
|
764 |
+
"content": "<SENTINEL_ID_27>",
|
765 |
+
"single_word": false,
|
766 |
+
"lstrip": false,
|
767 |
+
"rstrip": false,
|
768 |
+
"normalized": false,
|
769 |
+
"special": true
|
770 |
+
},
|
771 |
+
{
|
772 |
+
"id": 85,
|
773 |
+
"content": "<SENTINEL_ID_28>",
|
774 |
+
"single_word": false,
|
775 |
+
"lstrip": false,
|
776 |
+
"rstrip": false,
|
777 |
+
"normalized": false,
|
778 |
+
"special": true
|
779 |
+
},
|
780 |
+
{
|
781 |
+
"id": 86,
|
782 |
+
"content": "<SENTINEL_ID_29>",
|
783 |
+
"single_word": false,
|
784 |
+
"lstrip": false,
|
785 |
+
"rstrip": false,
|
786 |
+
"normalized": false,
|
787 |
+
"special": true
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"id": 87,
|
791 |
+
"content": "<SENTINEL_ID_30>",
|
792 |
+
"single_word": false,
|
793 |
+
"lstrip": false,
|
794 |
+
"rstrip": false,
|
795 |
+
"normalized": false,
|
796 |
+
"special": true
|
797 |
+
},
|
798 |
+
{
|
799 |
+
"id": 88,
|
800 |
+
"content": "<SENTINEL_ID_31>",
|
801 |
+
"single_word": false,
|
802 |
+
"lstrip": false,
|
803 |
+
"rstrip": false,
|
804 |
+
"normalized": false,
|
805 |
+
"special": true
|
806 |
+
},
|
807 |
+
{
|
808 |
+
"id": 89,
|
809 |
+
"content": "<SENTINEL_ID_32>",
|
810 |
+
"single_word": false,
|
811 |
+
"lstrip": false,
|
812 |
+
"rstrip": false,
|
813 |
+
"normalized": false,
|
814 |
+
"special": true
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"id": 90,
|
818 |
+
"content": "<SENTINEL_ID_33>",
|
819 |
+
"single_word": false,
|
820 |
+
"lstrip": false,
|
821 |
+
"rstrip": false,
|
822 |
+
"normalized": false,
|
823 |
+
"special": true
|
824 |
+
},
|
825 |
+
{
|
826 |
+
"id": 91,
|
827 |
+
"content": "<SENTINEL_ID_34>",
|
828 |
+
"single_word": false,
|
829 |
+
"lstrip": false,
|
830 |
+
"rstrip": false,
|
831 |
+
"normalized": false,
|
832 |
+
"special": true
|
833 |
+
},
|
834 |
+
{
|
835 |
+
"id": 92,
|
836 |
+
"content": "<SENTINEL_ID_35>",
|
837 |
+
"single_word": false,
|
838 |
+
"lstrip": false,
|
839 |
+
"rstrip": false,
|
840 |
+
"normalized": false,
|
841 |
+
"special": true
|
842 |
+
},
|
843 |
+
{
|
844 |
+
"id": 93,
|
845 |
+
"content": "<SENTINEL_ID_36>",
|
846 |
+
"single_word": false,
|
847 |
+
"lstrip": false,
|
848 |
+
"rstrip": false,
|
849 |
+
"normalized": false,
|
850 |
+
"special": true
|
851 |
+
},
|
852 |
+
{
|
853 |
+
"id": 94,
|
854 |
+
"content": "<SENTINEL_ID_37>",
|
855 |
+
"single_word": false,
|
856 |
+
"lstrip": false,
|
857 |
+
"rstrip": false,
|
858 |
+
"normalized": false,
|
859 |
+
"special": true
|
860 |
+
},
|
861 |
+
{
|
862 |
+
"id": 95,
|
863 |
+
"content": "<SENTINEL_ID_38>",
|
864 |
+
"single_word": false,
|
865 |
+
"lstrip": false,
|
866 |
+
"rstrip": false,
|
867 |
+
"normalized": false,
|
868 |
+
"special": true
|
869 |
+
},
|
870 |
+
{
|
871 |
+
"id": 96,
|
872 |
+
"content": "<SENTINEL_ID_39>",
|
873 |
+
"single_word": false,
|
874 |
+
"lstrip": false,
|
875 |
+
"rstrip": false,
|
876 |
+
"normalized": false,
|
877 |
+
"special": true
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"id": 97,
|
881 |
+
"content": "<SENTINEL_ID_40>",
|
882 |
+
"single_word": false,
|
883 |
+
"lstrip": false,
|
884 |
+
"rstrip": false,
|
885 |
+
"normalized": false,
|
886 |
+
"special": true
|
887 |
+
},
|
888 |
+
{
|
889 |
+
"id": 98,
|
890 |
+
"content": "<SENTINEL_ID_41>",
|
891 |
+
"single_word": false,
|
892 |
+
"lstrip": false,
|
893 |
+
"rstrip": false,
|
894 |
+
"normalized": false,
|
895 |
+
"special": true
|
896 |
+
},
|
897 |
+
{
|
898 |
+
"id": 99,
|
899 |
+
"content": "<SENTINEL_ID_42>",
|
900 |
+
"single_word": false,
|
901 |
+
"lstrip": false,
|
902 |
+
"rstrip": false,
|
903 |
+
"normalized": false,
|
904 |
+
"special": true
|
905 |
+
},
|
906 |
+
{
|
907 |
+
"id": 100,
|
908 |
+
"content": "<SENTINEL_ID_43>",
|
909 |
+
"single_word": false,
|
910 |
+
"lstrip": false,
|
911 |
+
"rstrip": false,
|
912 |
+
"normalized": false,
|
913 |
+
"special": true
|
914 |
+
},
|
915 |
+
{
|
916 |
+
"id": 101,
|
917 |
+
"content": "<SENTINEL_ID_44>",
|
918 |
+
"single_word": false,
|
919 |
+
"lstrip": false,
|
920 |
+
"rstrip": false,
|
921 |
+
"normalized": false,
|
922 |
+
"special": true
|
923 |
+
},
|
924 |
+
{
|
925 |
+
"id": 102,
|
926 |
+
"content": "<SENTINEL_ID_45>",
|
927 |
+
"single_word": false,
|
928 |
+
"lstrip": false,
|
929 |
+
"rstrip": false,
|
930 |
+
"normalized": false,
|
931 |
+
"special": true
|
932 |
+
},
|
933 |
+
{
|
934 |
+
"id": 103,
|
935 |
+
"content": "<SENTINEL_ID_46>",
|
936 |
+
"single_word": false,
|
937 |
+
"lstrip": false,
|
938 |
+
"rstrip": false,
|
939 |
+
"normalized": false,
|
940 |
+
"special": true
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"id": 104,
|
944 |
+
"content": "<SENTINEL_ID_47>",
|
945 |
+
"single_word": false,
|
946 |
+
"lstrip": false,
|
947 |
+
"rstrip": false,
|
948 |
+
"normalized": false,
|
949 |
+
"special": true
|
950 |
+
},
|
951 |
+
{
|
952 |
+
"id": 105,
|
953 |
+
"content": "<SENTINEL_ID_48>",
|
954 |
+
"single_word": false,
|
955 |
+
"lstrip": false,
|
956 |
+
"rstrip": false,
|
957 |
+
"normalized": false,
|
958 |
+
"special": true
|
959 |
+
},
|
960 |
+
{
|
961 |
+
"id": 106,
|
962 |
+
"content": "<SENTINEL_ID_49>",
|
963 |
+
"single_word": false,
|
964 |
+
"lstrip": false,
|
965 |
+
"rstrip": false,
|
966 |
+
"normalized": false,
|
967 |
+
"special": true
|
968 |
+
},
|
969 |
+
{
|
970 |
+
"id": 107,
|
971 |
+
"content": "<SENTINEL_ID_50>",
|
972 |
+
"single_word": false,
|
973 |
+
"lstrip": false,
|
974 |
+
"rstrip": false,
|
975 |
+
"normalized": false,
|
976 |
+
"special": true
|
977 |
+
},
|
978 |
+
{
|
979 |
+
"id": 108,
|
980 |
+
"content": "<SENTINEL_ID_51>",
|
981 |
+
"single_word": false,
|
982 |
+
"lstrip": false,
|
983 |
+
"rstrip": false,
|
984 |
+
"normalized": false,
|
985 |
+
"special": true
|
986 |
+
},
|
987 |
+
{
|
988 |
+
"id": 109,
|
989 |
+
"content": "<SENTINEL_ID_52>",
|
990 |
+
"single_word": false,
|
991 |
+
"lstrip": false,
|
992 |
+
"rstrip": false,
|
993 |
+
"normalized": false,
|
994 |
+
"special": true
|
995 |
+
},
|
996 |
+
{
|
997 |
+
"id": 110,
|
998 |
+
"content": "<SENTINEL_ID_53>",
|
999 |
+
"single_word": false,
|
1000 |
+
"lstrip": false,
|
1001 |
+
"rstrip": false,
|
1002 |
+
"normalized": false,
|
1003 |
+
"special": true
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"id": 111,
|
1007 |
+
"content": "<SENTINEL_ID_54>",
|
1008 |
+
"single_word": false,
|
1009 |
+
"lstrip": false,
|
1010 |
+
"rstrip": false,
|
1011 |
+
"normalized": false,
|
1012 |
+
"special": true
|
1013 |
+
},
|
1014 |
+
{
|
1015 |
+
"id": 112,
|
1016 |
+
"content": "<SENTINEL_ID_55>",
|
1017 |
+
"single_word": false,
|
1018 |
+
"lstrip": false,
|
1019 |
+
"rstrip": false,
|
1020 |
+
"normalized": false,
|
1021 |
+
"special": true
|
1022 |
+
},
|
1023 |
+
{
|
1024 |
+
"id": 113,
|
1025 |
+
"content": "<SENTINEL_ID_56>",
|
1026 |
+
"single_word": false,
|
1027 |
+
"lstrip": false,
|
1028 |
+
"rstrip": false,
|
1029 |
+
"normalized": false,
|
1030 |
+
"special": true
|
1031 |
+
},
|
1032 |
+
{
|
1033 |
+
"id": 114,
|
1034 |
+
"content": "<SENTINEL_ID_57>",
|
1035 |
+
"single_word": false,
|
1036 |
+
"lstrip": false,
|
1037 |
+
"rstrip": false,
|
1038 |
+
"normalized": false,
|
1039 |
+
"special": true
|
1040 |
+
},
|
1041 |
+
{
|
1042 |
+
"id": 115,
|
1043 |
+
"content": "<SENTINEL_ID_58>",
|
1044 |
+
"single_word": false,
|
1045 |
+
"lstrip": false,
|
1046 |
+
"rstrip": false,
|
1047 |
+
"normalized": false,
|
1048 |
+
"special": true
|
1049 |
+
},
|
1050 |
+
{
|
1051 |
+
"id": 116,
|
1052 |
+
"content": "<SENTINEL_ID_59>",
|
1053 |
+
"single_word": false,
|
1054 |
+
"lstrip": false,
|
1055 |
+
"rstrip": false,
|
1056 |
+
"normalized": false,
|
1057 |
+
"special": true
|
1058 |
+
},
|
1059 |
+
{
|
1060 |
+
"id": 117,
|
1061 |
+
"content": "<SENTINEL_ID_60>",
|
1062 |
+
"single_word": false,
|
1063 |
+
"lstrip": false,
|
1064 |
+
"rstrip": false,
|
1065 |
+
"normalized": false,
|
1066 |
+
"special": true
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"id": 118,
|
1070 |
+
"content": "<SENTINEL_ID_61>",
|
1071 |
+
"single_word": false,
|
1072 |
+
"lstrip": false,
|
1073 |
+
"rstrip": false,
|
1074 |
+
"normalized": false,
|
1075 |
+
"special": true
|
1076 |
+
},
|
1077 |
+
{
|
1078 |
+
"id": 119,
|
1079 |
+
"content": "<SENTINEL_ID_62>",
|
1080 |
+
"single_word": false,
|
1081 |
+
"lstrip": false,
|
1082 |
+
"rstrip": false,
|
1083 |
+
"normalized": false,
|
1084 |
+
"special": true
|
1085 |
+
},
|
1086 |
+
{
|
1087 |
+
"id": 120,
|
1088 |
+
"content": "<SENTINEL_ID_63>",
|
1089 |
+
"single_word": false,
|
1090 |
+
"lstrip": false,
|
1091 |
+
"rstrip": false,
|
1092 |
+
"normalized": false,
|
1093 |
+
"special": true
|
1094 |
+
},
|
1095 |
+
{
|
1096 |
+
"id": 121,
|
1097 |
+
"content": "<SENTINEL_ID_64>",
|
1098 |
+
"single_word": false,
|
1099 |
+
"lstrip": false,
|
1100 |
+
"rstrip": false,
|
1101 |
+
"normalized": false,
|
1102 |
+
"special": true
|
1103 |
+
},
|
1104 |
+
{
|
1105 |
+
"id": 122,
|
1106 |
+
"content": "<SENTINEL_ID_65>",
|
1107 |
+
"single_word": false,
|
1108 |
+
"lstrip": false,
|
1109 |
+
"rstrip": false,
|
1110 |
+
"normalized": false,
|
1111 |
+
"special": true
|
1112 |
+
},
|
1113 |
+
{
|
1114 |
+
"id": 123,
|
1115 |
+
"content": "<SENTINEL_ID_66>",
|
1116 |
+
"single_word": false,
|
1117 |
+
"lstrip": false,
|
1118 |
+
"rstrip": false,
|
1119 |
+
"normalized": false,
|
1120 |
+
"special": true
|
1121 |
+
},
|
1122 |
+
{
|
1123 |
+
"id": 124,
|
1124 |
+
"content": "<SENTINEL_ID_67>",
|
1125 |
+
"single_word": false,
|
1126 |
+
"lstrip": false,
|
1127 |
+
"rstrip": false,
|
1128 |
+
"normalized": false,
|
1129 |
+
"special": true
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"id": 125,
|
1133 |
+
"content": "<SENTINEL_ID_68>",
|
1134 |
+
"single_word": false,
|
1135 |
+
"lstrip": false,
|
1136 |
+
"rstrip": false,
|
1137 |
+
"normalized": false,
|
1138 |
+
"special": true
|
1139 |
+
},
|
1140 |
+
{
|
1141 |
+
"id": 126,
|
1142 |
+
"content": "<SENTINEL_ID_69>",
|
1143 |
+
"single_word": false,
|
1144 |
+
"lstrip": false,
|
1145 |
+
"rstrip": false,
|
1146 |
+
"normalized": false,
|
1147 |
+
"special": true
|
1148 |
+
},
|
1149 |
+
{
|
1150 |
+
"id": 127,
|
1151 |
+
"content": "<SENTINEL_ID_70>",
|
1152 |
+
"single_word": false,
|
1153 |
+
"lstrip": false,
|
1154 |
+
"rstrip": false,
|
1155 |
+
"normalized": false,
|
1156 |
+
"special": true
|
1157 |
+
},
|
1158 |
+
{
|
1159 |
+
"id": 128,
|
1160 |
+
"content": "<SENTINEL_ID_71>",
|
1161 |
+
"single_word": false,
|
1162 |
+
"lstrip": false,
|
1163 |
+
"rstrip": false,
|
1164 |
+
"normalized": false,
|
1165 |
+
"special": true
|
1166 |
+
},
|
1167 |
+
{
|
1168 |
+
"id": 129,
|
1169 |
+
"content": "<SENTINEL_ID_72>",
|
1170 |
+
"single_word": false,
|
1171 |
+
"lstrip": false,
|
1172 |
+
"rstrip": false,
|
1173 |
+
"normalized": false,
|
1174 |
+
"special": true
|
1175 |
+
},
|
1176 |
+
{
|
1177 |
+
"id": 130,
|
1178 |
+
"content": "<SENTINEL_ID_73>",
|
1179 |
+
"single_word": false,
|
1180 |
+
"lstrip": false,
|
1181 |
+
"rstrip": false,
|
1182 |
+
"normalized": false,
|
1183 |
+
"special": true
|
1184 |
+
},
|
1185 |
+
{
|
1186 |
+
"id": 131,
|
1187 |
+
"content": "<SENTINEL_ID_74>",
|
1188 |
+
"single_word": false,
|
1189 |
+
"lstrip": false,
|
1190 |
+
"rstrip": false,
|
1191 |
+
"normalized": false,
|
1192 |
+
"special": true
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"id": 132,
|
1196 |
+
"content": "<SENTINEL_ID_75>",
|
1197 |
+
"single_word": false,
|
1198 |
+
"lstrip": false,
|
1199 |
+
"rstrip": false,
|
1200 |
+
"normalized": false,
|
1201 |
+
"special": true
|
1202 |
+
},
|
1203 |
+
{
|
1204 |
+
"id": 133,
|
1205 |
+
"content": "<SENTINEL_ID_76>",
|
1206 |
+
"single_word": false,
|
1207 |
+
"lstrip": false,
|
1208 |
+
"rstrip": false,
|
1209 |
+
"normalized": false,
|
1210 |
+
"special": true
|
1211 |
+
},
|
1212 |
+
{
|
1213 |
+
"id": 134,
|
1214 |
+
"content": "<SENTINEL_ID_77>",
|
1215 |
+
"single_word": false,
|
1216 |
+
"lstrip": false,
|
1217 |
+
"rstrip": false,
|
1218 |
+
"normalized": false,
|
1219 |
+
"special": true
|
1220 |
+
},
|
1221 |
+
{
|
1222 |
+
"id": 135,
|
1223 |
+
"content": "<SENTINEL_ID_78>",
|
1224 |
+
"single_word": false,
|
1225 |
+
"lstrip": false,
|
1226 |
+
"rstrip": false,
|
1227 |
+
"normalized": false,
|
1228 |
+
"special": true
|
1229 |
+
},
|
1230 |
+
{
|
1231 |
+
"id": 136,
|
1232 |
+
"content": "<SENTINEL_ID_79>",
|
1233 |
+
"single_word": false,
|
1234 |
+
"lstrip": false,
|
1235 |
+
"rstrip": false,
|
1236 |
+
"normalized": false,
|
1237 |
+
"special": true
|
1238 |
+
},
|
1239 |
+
{
|
1240 |
+
"id": 137,
|
1241 |
+
"content": "<SENTINEL_ID_80>",
|
1242 |
+
"single_word": false,
|
1243 |
+
"lstrip": false,
|
1244 |
+
"rstrip": false,
|
1245 |
+
"normalized": false,
|
1246 |
+
"special": true
|
1247 |
+
},
|
1248 |
+
{
|
1249 |
+
"id": 138,
|
1250 |
+
"content": "<SENTINEL_ID_81>",
|
1251 |
+
"single_word": false,
|
1252 |
+
"lstrip": false,
|
1253 |
+
"rstrip": false,
|
1254 |
+
"normalized": false,
|
1255 |
+
"special": true
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"id": 139,
|
1259 |
+
"content": "<SENTINEL_ID_82>",
|
1260 |
+
"single_word": false,
|
1261 |
+
"lstrip": false,
|
1262 |
+
"rstrip": false,
|
1263 |
+
"normalized": false,
|
1264 |
+
"special": true
|
1265 |
+
},
|
1266 |
+
{
|
1267 |
+
"id": 140,
|
1268 |
+
"content": "<SENTINEL_ID_83>",
|
1269 |
+
"single_word": false,
|
1270 |
+
"lstrip": false,
|
1271 |
+
"rstrip": false,
|
1272 |
+
"normalized": false,
|
1273 |
+
"special": true
|
1274 |
+
},
|
1275 |
+
{
|
1276 |
+
"id": 141,
|
1277 |
+
"content": "<SENTINEL_ID_84>",
|
1278 |
+
"single_word": false,
|
1279 |
+
"lstrip": false,
|
1280 |
+
"rstrip": false,
|
1281 |
+
"normalized": false,
|
1282 |
+
"special": true
|
1283 |
+
},
|
1284 |
+
{
|
1285 |
+
"id": 142,
|
1286 |
+
"content": "<SENTINEL_ID_85>",
|
1287 |
+
"single_word": false,
|
1288 |
+
"lstrip": false,
|
1289 |
+
"rstrip": false,
|
1290 |
+
"normalized": false,
|
1291 |
+
"special": true
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"id": 143,
|
1295 |
+
"content": "<SENTINEL_ID_86>",
|
1296 |
+
"single_word": false,
|
1297 |
+
"lstrip": false,
|
1298 |
+
"rstrip": false,
|
1299 |
+
"normalized": false,
|
1300 |
+
"special": true
|
1301 |
+
},
|
1302 |
+
{
|
1303 |
+
"id": 144,
|
1304 |
+
"content": "<SENTINEL_ID_87>",
|
1305 |
+
"single_word": false,
|
1306 |
+
"lstrip": false,
|
1307 |
+
"rstrip": false,
|
1308 |
+
"normalized": false,
|
1309 |
+
"special": true
|
1310 |
+
},
|
1311 |
+
{
|
1312 |
+
"id": 145,
|
1313 |
+
"content": "<SENTINEL_ID_88>",
|
1314 |
+
"single_word": false,
|
1315 |
+
"lstrip": false,
|
1316 |
+
"rstrip": false,
|
1317 |
+
"normalized": false,
|
1318 |
+
"special": true
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"id": 146,
|
1322 |
+
"content": "<SENTINEL_ID_89>",
|
1323 |
+
"single_word": false,
|
1324 |
+
"lstrip": false,
|
1325 |
+
"rstrip": false,
|
1326 |
+
"normalized": false,
|
1327 |
+
"special": true
|
1328 |
+
},
|
1329 |
+
{
|
1330 |
+
"id": 147,
|
1331 |
+
"content": "<SENTINEL_ID_90>",
|
1332 |
+
"single_word": false,
|
1333 |
+
"lstrip": false,
|
1334 |
+
"rstrip": false,
|
1335 |
+
"normalized": false,
|
1336 |
+
"special": true
|
1337 |
+
},
|
1338 |
+
{
|
1339 |
+
"id": 148,
|
1340 |
+
"content": "<SENTINEL_ID_91>",
|
1341 |
+
"single_word": false,
|
1342 |
+
"lstrip": false,
|
1343 |
+
"rstrip": false,
|
1344 |
+
"normalized": false,
|
1345 |
+
"special": true
|
1346 |
+
},
|
1347 |
+
{
|
1348 |
+
"id": 149,
|
1349 |
+
"content": "<SENTINEL_ID_92>",
|
1350 |
+
"single_word": false,
|
1351 |
+
"lstrip": false,
|
1352 |
+
"rstrip": false,
|
1353 |
+
"normalized": false,
|
1354 |
+
"special": true
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"id": 150,
|
1358 |
+
"content": "<SENTINEL_ID_93>",
|
1359 |
+
"single_word": false,
|
1360 |
+
"lstrip": false,
|
1361 |
+
"rstrip": false,
|
1362 |
+
"normalized": false,
|
1363 |
+
"special": true
|
1364 |
+
},
|
1365 |
+
{
|
1366 |
+
"id": 151,
|
1367 |
+
"content": "<SENTINEL_ID_94>",
|
1368 |
+
"single_word": false,
|
1369 |
+
"lstrip": false,
|
1370 |
+
"rstrip": false,
|
1371 |
+
"normalized": false,
|
1372 |
+
"special": true
|
1373 |
+
},
|
1374 |
+
{
|
1375 |
+
"id": 152,
|
1376 |
+
"content": "<SENTINEL_ID_95>",
|
1377 |
+
"single_word": false,
|
1378 |
+
"lstrip": false,
|
1379 |
+
"rstrip": false,
|
1380 |
+
"normalized": false,
|
1381 |
+
"special": true
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"id": 153,
|
1385 |
+
"content": "<SENTINEL_ID_96>",
|
1386 |
+
"single_word": false,
|
1387 |
+
"lstrip": false,
|
1388 |
+
"rstrip": false,
|
1389 |
+
"normalized": false,
|
1390 |
+
"special": true
|
1391 |
+
},
|
1392 |
+
{
|
1393 |
+
"id": 154,
|
1394 |
+
"content": "<SENTINEL_ID_97>",
|
1395 |
+
"single_word": false,
|
1396 |
+
"lstrip": false,
|
1397 |
+
"rstrip": false,
|
1398 |
+
"normalized": false,
|
1399 |
+
"special": true
|
1400 |
+
},
|
1401 |
+
{
|
1402 |
+
"id": 155,
|
1403 |
+
"content": "<SENTINEL_ID_98>",
|
1404 |
+
"single_word": false,
|
1405 |
+
"lstrip": false,
|
1406 |
+
"rstrip": false,
|
1407 |
+
"normalized": false,
|
1408 |
+
"special": true
|
1409 |
+
},
|
1410 |
+
{
|
1411 |
+
"id": 156,
|
1412 |
+
"content": "<SENTINEL_ID_99>",
|
1413 |
+
"single_word": false,
|
1414 |
+
"lstrip": false,
|
1415 |
+
"rstrip": false,
|
1416 |
+
"normalized": false,
|
1417 |
+
"special": true
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"id": 157,
|
1421 |
+
"content": "<SENTINEL_ID_100>",
|
1422 |
+
"single_word": false,
|
1423 |
+
"lstrip": false,
|
1424 |
+
"rstrip": false,
|
1425 |
+
"normalized": false,
|
1426 |
+
"special": true
|
1427 |
+
},
|
1428 |
+
{
|
1429 |
+
"id": 158,
|
1430 |
+
"content": "<SENTINEL_ID_101>",
|
1431 |
+
"single_word": false,
|
1432 |
+
"lstrip": false,
|
1433 |
+
"rstrip": false,
|
1434 |
+
"normalized": false,
|
1435 |
+
"special": true
|
1436 |
+
},
|
1437 |
+
{
|
1438 |
+
"id": 159,
|
1439 |
+
"content": "<SENTINEL_ID_102>",
|
1440 |
+
"single_word": false,
|
1441 |
+
"lstrip": false,
|
1442 |
+
"rstrip": false,
|
1443 |
+
"normalized": false,
|
1444 |
+
"special": true
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"id": 160,
|
1448 |
+
"content": "<SENTINEL_ID_103>",
|
1449 |
+
"single_word": false,
|
1450 |
+
"lstrip": false,
|
1451 |
+
"rstrip": false,
|
1452 |
+
"normalized": false,
|
1453 |
+
"special": true
|
1454 |
+
},
|
1455 |
+
{
|
1456 |
+
"id": 161,
|
1457 |
+
"content": "<SENTINEL_ID_104>",
|
1458 |
+
"single_word": false,
|
1459 |
+
"lstrip": false,
|
1460 |
+
"rstrip": false,
|
1461 |
+
"normalized": false,
|
1462 |
+
"special": true
|
1463 |
+
},
|
1464 |
+
{
|
1465 |
+
"id": 162,
|
1466 |
+
"content": "<SENTINEL_ID_105>",
|
1467 |
+
"single_word": false,
|
1468 |
+
"lstrip": false,
|
1469 |
+
"rstrip": false,
|
1470 |
+
"normalized": false,
|
1471 |
+
"special": true
|
1472 |
+
},
|
1473 |
+
{
|
1474 |
+
"id": 163,
|
1475 |
+
"content": "<SENTINEL_ID_106>",
|
1476 |
+
"single_word": false,
|
1477 |
+
"lstrip": false,
|
1478 |
+
"rstrip": false,
|
1479 |
+
"normalized": false,
|
1480 |
+
"special": true
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"id": 164,
|
1484 |
+
"content": "<SENTINEL_ID_107>",
|
1485 |
+
"single_word": false,
|
1486 |
+
"lstrip": false,
|
1487 |
+
"rstrip": false,
|
1488 |
+
"normalized": false,
|
1489 |
+
"special": true
|
1490 |
+
},
|
1491 |
+
{
|
1492 |
+
"id": 165,
|
1493 |
+
"content": "<SENTINEL_ID_108>",
|
1494 |
+
"single_word": false,
|
1495 |
+
"lstrip": false,
|
1496 |
+
"rstrip": false,
|
1497 |
+
"normalized": false,
|
1498 |
+
"special": true
|
1499 |
+
},
|
1500 |
+
{
|
1501 |
+
"id": 166,
|
1502 |
+
"content": "<SENTINEL_ID_109>",
|
1503 |
+
"single_word": false,
|
1504 |
+
"lstrip": false,
|
1505 |
+
"rstrip": false,
|
1506 |
+
"normalized": false,
|
1507 |
+
"special": true
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"id": 167,
|
1511 |
+
"content": "<SENTINEL_ID_110>",
|
1512 |
+
"single_word": false,
|
1513 |
+
"lstrip": false,
|
1514 |
+
"rstrip": false,
|
1515 |
+
"normalized": false,
|
1516 |
+
"special": true
|
1517 |
+
},
|
1518 |
+
{
|
1519 |
+
"id": 168,
|
1520 |
+
"content": "<SENTINEL_ID_111>",
|
1521 |
+
"single_word": false,
|
1522 |
+
"lstrip": false,
|
1523 |
+
"rstrip": false,
|
1524 |
+
"normalized": false,
|
1525 |
+
"special": true
|
1526 |
+
},
|
1527 |
+
{
|
1528 |
+
"id": 169,
|
1529 |
+
"content": "<SENTINEL_ID_112>",
|
1530 |
+
"single_word": false,
|
1531 |
+
"lstrip": false,
|
1532 |
+
"rstrip": false,
|
1533 |
+
"normalized": false,
|
1534 |
+
"special": true
|
1535 |
+
},
|
1536 |
+
{
|
1537 |
+
"id": 170,
|
1538 |
+
"content": "<SENTINEL_ID_113>",
|
1539 |
+
"single_word": false,
|
1540 |
+
"lstrip": false,
|
1541 |
+
"rstrip": false,
|
1542 |
+
"normalized": false,
|
1543 |
+
"special": true
|
1544 |
+
},
|
1545 |
+
{
|
1546 |
+
"id": 171,
|
1547 |
+
"content": "<SENTINEL_ID_114>",
|
1548 |
+
"single_word": false,
|
1549 |
+
"lstrip": false,
|
1550 |
+
"rstrip": false,
|
1551 |
+
"normalized": false,
|
1552 |
+
"special": true
|
1553 |
+
},
|
1554 |
+
{
|
1555 |
+
"id": 172,
|
1556 |
+
"content": "<SENTINEL_ID_115>",
|
1557 |
+
"single_word": false,
|
1558 |
+
"lstrip": false,
|
1559 |
+
"rstrip": false,
|
1560 |
+
"normalized": false,
|
1561 |
+
"special": true
|
1562 |
+
},
|
1563 |
+
{
|
1564 |
+
"id": 173,
|
1565 |
+
"content": "<SENTINEL_ID_116>",
|
1566 |
+
"single_word": false,
|
1567 |
+
"lstrip": false,
|
1568 |
+
"rstrip": false,
|
1569 |
+
"normalized": false,
|
1570 |
+
"special": true
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"id": 174,
|
1574 |
+
"content": "<SENTINEL_ID_117>",
|
1575 |
+
"single_word": false,
|
1576 |
+
"lstrip": false,
|
1577 |
+
"rstrip": false,
|
1578 |
+
"normalized": false,
|
1579 |
+
"special": true
|
1580 |
+
},
|
1581 |
+
{
|
1582 |
+
"id": 175,
|
1583 |
+
"content": "<SENTINEL_ID_118>",
|
1584 |
+
"single_word": false,
|
1585 |
+
"lstrip": false,
|
1586 |
+
"rstrip": false,
|
1587 |
+
"normalized": false,
|
1588 |
+
"special": true
|
1589 |
+
},
|
1590 |
+
{
|
1591 |
+
"id": 176,
|
1592 |
+
"content": "<SENTINEL_ID_119>",
|
1593 |
+
"single_word": false,
|
1594 |
+
"lstrip": false,
|
1595 |
+
"rstrip": false,
|
1596 |
+
"normalized": false,
|
1597 |
+
"special": true
|
1598 |
+
},
|
1599 |
+
{
|
1600 |
+
"id": 177,
|
1601 |
+
"content": "<SENTINEL_ID_120>",
|
1602 |
+
"single_word": false,
|
1603 |
+
"lstrip": false,
|
1604 |
+
"rstrip": false,
|
1605 |
+
"normalized": false,
|
1606 |
+
"special": true
|
1607 |
+
},
|
1608 |
+
{
|
1609 |
+
"id": 178,
|
1610 |
+
"content": "<SENTINEL_ID_121>",
|
1611 |
+
"single_word": false,
|
1612 |
+
"lstrip": false,
|
1613 |
+
"rstrip": false,
|
1614 |
+
"normalized": false,
|
1615 |
+
"special": true
|
1616 |
+
},
|
1617 |
+
{
|
1618 |
+
"id": 179,
|
1619 |
+
"content": "<SENTINEL_ID_122>",
|
1620 |
+
"single_word": false,
|
1621 |
+
"lstrip": false,
|
1622 |
+
"rstrip": false,
|
1623 |
+
"normalized": false,
|
1624 |
+
"special": true
|
1625 |
+
},
|
1626 |
+
{
|
1627 |
+
"id": 180,
|
1628 |
+
"content": "<SENTINEL_ID_123>",
|
1629 |
+
"single_word": false,
|
1630 |
+
"lstrip": false,
|
1631 |
+
"rstrip": false,
|
1632 |
+
"normalized": false,
|
1633 |
+
"special": true
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"id": 181,
|
1637 |
+
"content": "<SENTINEL_ID_124>",
|
1638 |
+
"single_word": false,
|
1639 |
+
"lstrip": false,
|
1640 |
+
"rstrip": false,
|
1641 |
+
"normalized": false,
|
1642 |
+
"special": true
|
1643 |
+
},
|
1644 |
+
{
|
1645 |
+
"id": 182,
|
1646 |
+
"content": "<SENTINEL_ID_125>",
|
1647 |
+
"single_word": false,
|
1648 |
+
"lstrip": false,
|
1649 |
+
"rstrip": false,
|
1650 |
+
"normalized": false,
|
1651 |
+
"special": true
|
1652 |
+
},
|
1653 |
+
{
|
1654 |
+
"id": 183,
|
1655 |
+
"content": "<SENTINEL_ID_126>",
|
1656 |
+
"single_word": false,
|
1657 |
+
"lstrip": false,
|
1658 |
+
"rstrip": false,
|
1659 |
+
"normalized": false,
|
1660 |
+
"special": true
|
1661 |
+
},
|
1662 |
+
{
|
1663 |
+
"id": 184,
|
1664 |
+
"content": "<SENTINEL_ID_127>",
|
1665 |
+
"single_word": false,
|
1666 |
+
"lstrip": false,
|
1667 |
+
"rstrip": false,
|
1668 |
+
"normalized": false,
|
1669 |
+
"special": true
|
1670 |
+
},
|
1671 |
+
{
|
1672 |
+
"id": 185,
|
1673 |
+
"content": "<SENTINEL_ID_128>",
|
1674 |
+
"single_word": false,
|
1675 |
+
"lstrip": false,
|
1676 |
+
"rstrip": false,
|
1677 |
+
"normalized": false,
|
1678 |
+
"special": true
|
1679 |
+
},
|
1680 |
+
{
|
1681 |
+
"id": 186,
|
1682 |
+
"content": "<SENTINEL_ID_129>",
|
1683 |
+
"single_word": false,
|
1684 |
+
"lstrip": false,
|
1685 |
+
"rstrip": false,
|
1686 |
+
"normalized": false,
|
1687 |
+
"special": true
|
1688 |
+
},
|
1689 |
+
{
|
1690 |
+
"id": 187,
|
1691 |
+
"content": "<SENTINEL_ID_130>",
|
1692 |
+
"single_word": false,
|
1693 |
+
"lstrip": false,
|
1694 |
+
"rstrip": false,
|
1695 |
+
"normalized": false,
|
1696 |
+
"special": true
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"id": 188,
|
1700 |
+
"content": "<SENTINEL_ID_131>",
|
1701 |
+
"single_word": false,
|
1702 |
+
"lstrip": false,
|
1703 |
+
"rstrip": false,
|
1704 |
+
"normalized": false,
|
1705 |
+
"special": true
|
1706 |
+
},
|
1707 |
+
{
|
1708 |
+
"id": 189,
|
1709 |
+
"content": "<SENTINEL_ID_132>",
|
1710 |
+
"single_word": false,
|
1711 |
+
"lstrip": false,
|
1712 |
+
"rstrip": false,
|
1713 |
+
"normalized": false,
|
1714 |
+
"special": true
|
1715 |
+
},
|
1716 |
+
{
|
1717 |
+
"id": 190,
|
1718 |
+
"content": "<SENTINEL_ID_133>",
|
1719 |
+
"single_word": false,
|
1720 |
+
"lstrip": false,
|
1721 |
+
"rstrip": false,
|
1722 |
+
"normalized": false,
|
1723 |
+
"special": true
|
1724 |
+
},
|
1725 |
+
{
|
1726 |
+
"id": 191,
|
1727 |
+
"content": "<SENTINEL_ID_134>",
|
1728 |
+
"single_word": false,
|
1729 |
+
"lstrip": false,
|
1730 |
+
"rstrip": false,
|
1731 |
+
"normalized": false,
|
1732 |
+
"special": true
|
1733 |
+
},
|
1734 |
+
{
|
1735 |
+
"id": 192,
|
1736 |
+
"content": "<SENTINEL_ID_135>",
|
1737 |
+
"single_word": false,
|
1738 |
+
"lstrip": false,
|
1739 |
+
"rstrip": false,
|
1740 |
+
"normalized": false,
|
1741 |
+
"special": true
|
1742 |
+
},
|
1743 |
+
{
|
1744 |
+
"id": 193,
|
1745 |
+
"content": "<SENTINEL_ID_136>",
|
1746 |
+
"single_word": false,
|
1747 |
+
"lstrip": false,
|
1748 |
+
"rstrip": false,
|
1749 |
+
"normalized": false,
|
1750 |
+
"special": true
|
1751 |
+
},
|
1752 |
+
{
|
1753 |
+
"id": 194,
|
1754 |
+
"content": "<SENTINEL_ID_137>",
|
1755 |
+
"single_word": false,
|
1756 |
+
"lstrip": false,
|
1757 |
+
"rstrip": false,
|
1758 |
+
"normalized": false,
|
1759 |
+
"special": true
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"id": 195,
|
1763 |
+
"content": "<SENTINEL_ID_138>",
|
1764 |
+
"single_word": false,
|
1765 |
+
"lstrip": false,
|
1766 |
+
"rstrip": false,
|
1767 |
+
"normalized": false,
|
1768 |
+
"special": true
|
1769 |
+
},
|
1770 |
+
{
|
1771 |
+
"id": 196,
|
1772 |
+
"content": "<SENTINEL_ID_139>",
|
1773 |
+
"single_word": false,
|
1774 |
+
"lstrip": false,
|
1775 |
+
"rstrip": false,
|
1776 |
+
"normalized": false,
|
1777 |
+
"special": true
|
1778 |
+
},
|
1779 |
+
{
|
1780 |
+
"id": 197,
|
1781 |
+
"content": "<SENTINEL_ID_140>",
|
1782 |
+
"single_word": false,
|
1783 |
+
"lstrip": false,
|
1784 |
+
"rstrip": false,
|
1785 |
+
"normalized": false,
|
1786 |
+
"special": true
|
1787 |
+
},
|
1788 |
+
{
|
1789 |
+
"id": 198,
|
1790 |
+
"content": "<SENTINEL_ID_141>",
|
1791 |
+
"single_word": false,
|
1792 |
+
"lstrip": false,
|
1793 |
+
"rstrip": false,
|
1794 |
+
"normalized": false,
|
1795 |
+
"special": true
|
1796 |
+
},
|
1797 |
+
{
|
1798 |
+
"id": 199,
|
1799 |
+
"content": "<SENTINEL_ID_142>",
|
1800 |
+
"single_word": false,
|
1801 |
+
"lstrip": false,
|
1802 |
+
"rstrip": false,
|
1803 |
+
"normalized": false,
|
1804 |
+
"special": true
|
1805 |
+
},
|
1806 |
+
{
|
1807 |
+
"id": 200,
|
1808 |
+
"content": "<SENTINEL_ID_143>",
|
1809 |
+
"single_word": false,
|
1810 |
+
"lstrip": false,
|
1811 |
+
"rstrip": false,
|
1812 |
+
"normalized": false,
|
1813 |
+
"special": true
|
1814 |
+
},
|
1815 |
+
{
|
1816 |
+
"id": 201,
|
1817 |
+
"content": "<SENTINEL_ID_144>",
|
1818 |
+
"single_word": false,
|
1819 |
+
"lstrip": false,
|
1820 |
+
"rstrip": false,
|
1821 |
+
"normalized": false,
|
1822 |
+
"special": true
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"id": 202,
|
1826 |
+
"content": "<SENTINEL_ID_145>",
|
1827 |
+
"single_word": false,
|
1828 |
+
"lstrip": false,
|
1829 |
+
"rstrip": false,
|
1830 |
+
"normalized": false,
|
1831 |
+
"special": true
|
1832 |
+
},
|
1833 |
+
{
|
1834 |
+
"id": 203,
|
1835 |
+
"content": "<SENTINEL_ID_146>",
|
1836 |
+
"single_word": false,
|
1837 |
+
"lstrip": false,
|
1838 |
+
"rstrip": false,
|
1839 |
+
"normalized": false,
|
1840 |
+
"special": true
|
1841 |
+
},
|
1842 |
+
{
|
1843 |
+
"id": 204,
|
1844 |
+
"content": "<SENTINEL_ID_147>",
|
1845 |
+
"single_word": false,
|
1846 |
+
"lstrip": false,
|
1847 |
+
"rstrip": false,
|
1848 |
+
"normalized": false,
|
1849 |
+
"special": true
|
1850 |
+
},
|
1851 |
+
{
|
1852 |
+
"id": 205,
|
1853 |
+
"content": "<SENTINEL_ID_148>",
|
1854 |
+
"single_word": false,
|
1855 |
+
"lstrip": false,
|
1856 |
+
"rstrip": false,
|
1857 |
+
"normalized": false,
|
1858 |
+
"special": true
|
1859 |
+
},
|
1860 |
+
{
|
1861 |
+
"id": 206,
|
1862 |
+
"content": "<SENTINEL_ID_149>",
|
1863 |
+
"single_word": false,
|
1864 |
+
"lstrip": false,
|
1865 |
+
"rstrip": false,
|
1866 |
+
"normalized": false,
|
1867 |
+
"special": true
|
1868 |
+
},
|
1869 |
+
{
|
1870 |
+
"id": 207,
|
1871 |
+
"content": "<SENTINEL_ID_150>",
|
1872 |
+
"single_word": false,
|
1873 |
+
"lstrip": false,
|
1874 |
+
"rstrip": false,
|
1875 |
+
"normalized": false,
|
1876 |
+
"special": true
|
1877 |
+
},
|
1878 |
+
{
|
1879 |
+
"id": 208,
|
1880 |
+
"content": "<SENTINEL_ID_151>",
|
1881 |
+
"single_word": false,
|
1882 |
+
"lstrip": false,
|
1883 |
+
"rstrip": false,
|
1884 |
+
"normalized": false,
|
1885 |
+
"special": true
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"id": 209,
|
1889 |
+
"content": "<SENTINEL_ID_152>",
|
1890 |
+
"single_word": false,
|
1891 |
+
"lstrip": false,
|
1892 |
+
"rstrip": false,
|
1893 |
+
"normalized": false,
|
1894 |
+
"special": true
|
1895 |
+
},
|
1896 |
+
{
|
1897 |
+
"id": 210,
|
1898 |
+
"content": "<SENTINEL_ID_153>",
|
1899 |
+
"single_word": false,
|
1900 |
+
"lstrip": false,
|
1901 |
+
"rstrip": false,
|
1902 |
+
"normalized": false,
|
1903 |
+
"special": true
|
1904 |
+
},
|
1905 |
+
{
|
1906 |
+
"id": 211,
|
1907 |
+
"content": "<SENTINEL_ID_154>",
|
1908 |
+
"single_word": false,
|
1909 |
+
"lstrip": false,
|
1910 |
+
"rstrip": false,
|
1911 |
+
"normalized": false,
|
1912 |
+
"special": true
|
1913 |
+
},
|
1914 |
+
{
|
1915 |
+
"id": 212,
|
1916 |
+
"content": "<SENTINEL_ID_155>",
|
1917 |
+
"single_word": false,
|
1918 |
+
"lstrip": false,
|
1919 |
+
"rstrip": false,
|
1920 |
+
"normalized": false,
|
1921 |
+
"special": true
|
1922 |
+
},
|
1923 |
+
{
|
1924 |
+
"id": 213,
|
1925 |
+
"content": "<SENTINEL_ID_156>",
|
1926 |
+
"single_word": false,
|
1927 |
+
"lstrip": false,
|
1928 |
+
"rstrip": false,
|
1929 |
+
"normalized": false,
|
1930 |
+
"special": true
|
1931 |
+
},
|
1932 |
+
{
|
1933 |
+
"id": 214,
|
1934 |
+
"content": "<SENTINEL_ID_157>",
|
1935 |
+
"single_word": false,
|
1936 |
+
"lstrip": false,
|
1937 |
+
"rstrip": false,
|
1938 |
+
"normalized": false,
|
1939 |
+
"special": true
|
1940 |
+
},
|
1941 |
+
{
|
1942 |
+
"id": 215,
|
1943 |
+
"content": "<SENTINEL_ID_158>",
|
1944 |
+
"single_word": false,
|
1945 |
+
"lstrip": false,
|
1946 |
+
"rstrip": false,
|
1947 |
+
"normalized": false,
|
1948 |
+
"special": true
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"id": 216,
|
1952 |
+
"content": "<SENTINEL_ID_159>",
|
1953 |
+
"single_word": false,
|
1954 |
+
"lstrip": false,
|
1955 |
+
"rstrip": false,
|
1956 |
+
"normalized": false,
|
1957 |
+
"special": true
|
1958 |
+
},
|
1959 |
+
{
|
1960 |
+
"id": 217,
|
1961 |
+
"content": "<SENTINEL_ID_160>",
|
1962 |
+
"single_word": false,
|
1963 |
+
"lstrip": false,
|
1964 |
+
"rstrip": false,
|
1965 |
+
"normalized": false,
|
1966 |
+
"special": true
|
1967 |
+
},
|
1968 |
+
{
|
1969 |
+
"id": 218,
|
1970 |
+
"content": "<SENTINEL_ID_161>",
|
1971 |
+
"single_word": false,
|
1972 |
+
"lstrip": false,
|
1973 |
+
"rstrip": false,
|
1974 |
+
"normalized": false,
|
1975 |
+
"special": true
|
1976 |
+
},
|
1977 |
+
{
|
1978 |
+
"id": 219,
|
1979 |
+
"content": "<SENTINEL_ID_162>",
|
1980 |
+
"single_word": false,
|
1981 |
+
"lstrip": false,
|
1982 |
+
"rstrip": false,
|
1983 |
+
"normalized": false,
|
1984 |
+
"special": true
|
1985 |
+
},
|
1986 |
+
{
|
1987 |
+
"id": 220,
|
1988 |
+
"content": "<SENTINEL_ID_163>",
|
1989 |
+
"single_word": false,
|
1990 |
+
"lstrip": false,
|
1991 |
+
"rstrip": false,
|
1992 |
+
"normalized": false,
|
1993 |
+
"special": true
|
1994 |
+
},
|
1995 |
+
{
|
1996 |
+
"id": 221,
|
1997 |
+
"content": "<SENTINEL_ID_164>",
|
1998 |
+
"single_word": false,
|
1999 |
+
"lstrip": false,
|
2000 |
+
"rstrip": false,
|
2001 |
+
"normalized": false,
|
2002 |
+
"special": true
|
2003 |
+
},
|
2004 |
+
{
|
2005 |
+
"id": 222,
|
2006 |
+
"content": "<SENTINEL_ID_165>",
|
2007 |
+
"single_word": false,
|
2008 |
+
"lstrip": false,
|
2009 |
+
"rstrip": false,
|
2010 |
+
"normalized": false,
|
2011 |
+
"special": true
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"id": 223,
|
2015 |
+
"content": "<SENTINEL_ID_166>",
|
2016 |
+
"single_word": false,
|
2017 |
+
"lstrip": false,
|
2018 |
+
"rstrip": false,
|
2019 |
+
"normalized": false,
|
2020 |
+
"special": true
|
2021 |
+
},
|
2022 |
+
{
|
2023 |
+
"id": 224,
|
2024 |
+
"content": "<SENTINEL_ID_167>",
|
2025 |
+
"single_word": false,
|
2026 |
+
"lstrip": false,
|
2027 |
+
"rstrip": false,
|
2028 |
+
"normalized": false,
|
2029 |
+
"special": true
|
2030 |
+
},
|
2031 |
+
{
|
2032 |
+
"id": 225,
|
2033 |
+
"content": "<SENTINEL_ID_168>",
|
2034 |
+
"single_word": false,
|
2035 |
+
"lstrip": false,
|
2036 |
+
"rstrip": false,
|
2037 |
+
"normalized": false,
|
2038 |
+
"special": true
|
2039 |
+
},
|
2040 |
+
{
|
2041 |
+
"id": 226,
|
2042 |
+
"content": "<SENTINEL_ID_169>",
|
2043 |
+
"single_word": false,
|
2044 |
+
"lstrip": false,
|
2045 |
+
"rstrip": false,
|
2046 |
+
"normalized": false,
|
2047 |
+
"special": true
|
2048 |
+
},
|
2049 |
+
{
|
2050 |
+
"id": 227,
|
2051 |
+
"content": "<SENTINEL_ID_170>",
|
2052 |
+
"single_word": false,
|
2053 |
+
"lstrip": false,
|
2054 |
+
"rstrip": false,
|
2055 |
+
"normalized": false,
|
2056 |
+
"special": true
|
2057 |
+
},
|
2058 |
+
{
|
2059 |
+
"id": 228,
|
2060 |
+
"content": "<SENTINEL_ID_171>",
|
2061 |
+
"single_word": false,
|
2062 |
+
"lstrip": false,
|
2063 |
+
"rstrip": false,
|
2064 |
+
"normalized": false,
|
2065 |
+
"special": true
|
2066 |
+
},
|
2067 |
+
{
|
2068 |
+
"id": 229,
|
2069 |
+
"content": "<SENTINEL_ID_172>",
|
2070 |
+
"single_word": false,
|
2071 |
+
"lstrip": false,
|
2072 |
+
"rstrip": false,
|
2073 |
+
"normalized": false,
|
2074 |
+
"special": true
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"id": 230,
|
2078 |
+
"content": "<SENTINEL_ID_173>",
|
2079 |
+
"single_word": false,
|
2080 |
+
"lstrip": false,
|
2081 |
+
"rstrip": false,
|
2082 |
+
"normalized": false,
|
2083 |
+
"special": true
|
2084 |
+
},
|
2085 |
+
{
|
2086 |
+
"id": 231,
|
2087 |
+
"content": "<SENTINEL_ID_174>",
|
2088 |
+
"single_word": false,
|
2089 |
+
"lstrip": false,
|
2090 |
+
"rstrip": false,
|
2091 |
+
"normalized": false,
|
2092 |
+
"special": true
|
2093 |
+
},
|
2094 |
+
{
|
2095 |
+
"id": 232,
|
2096 |
+
"content": "<SENTINEL_ID_175>",
|
2097 |
+
"single_word": false,
|
2098 |
+
"lstrip": false,
|
2099 |
+
"rstrip": false,
|
2100 |
+
"normalized": false,
|
2101 |
+
"special": true
|
2102 |
+
},
|
2103 |
+
{
|
2104 |
+
"id": 233,
|
2105 |
+
"content": "<SENTINEL_ID_176>",
|
2106 |
+
"single_word": false,
|
2107 |
+
"lstrip": false,
|
2108 |
+
"rstrip": false,
|
2109 |
+
"normalized": false,
|
2110 |
+
"special": true
|
2111 |
+
},
|
2112 |
+
{
|
2113 |
+
"id": 234,
|
2114 |
+
"content": "<SENTINEL_ID_177>",
|
2115 |
+
"single_word": false,
|
2116 |
+
"lstrip": false,
|
2117 |
+
"rstrip": false,
|
2118 |
+
"normalized": false,
|
2119 |
+
"special": true
|
2120 |
+
},
|
2121 |
+
{
|
2122 |
+
"id": 235,
|
2123 |
+
"content": "<SENTINEL_ID_178>",
|
2124 |
+
"single_word": false,
|
2125 |
+
"lstrip": false,
|
2126 |
+
"rstrip": false,
|
2127 |
+
"normalized": false,
|
2128 |
+
"special": true
|
2129 |
+
},
|
2130 |
+
{
|
2131 |
+
"id": 236,
|
2132 |
+
"content": "<SENTINEL_ID_179>",
|
2133 |
+
"single_word": false,
|
2134 |
+
"lstrip": false,
|
2135 |
+
"rstrip": false,
|
2136 |
+
"normalized": false,
|
2137 |
+
"special": true
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"id": 237,
|
2141 |
+
"content": "<SENTINEL_ID_180>",
|
2142 |
+
"single_word": false,
|
2143 |
+
"lstrip": false,
|
2144 |
+
"rstrip": false,
|
2145 |
+
"normalized": false,
|
2146 |
+
"special": true
|
2147 |
+
},
|
2148 |
+
{
|
2149 |
+
"id": 238,
|
2150 |
+
"content": "<SENTINEL_ID_181>",
|
2151 |
+
"single_word": false,
|
2152 |
+
"lstrip": false,
|
2153 |
+
"rstrip": false,
|
2154 |
+
"normalized": false,
|
2155 |
+
"special": true
|
2156 |
+
},
|
2157 |
+
{
|
2158 |
+
"id": 239,
|
2159 |
+
"content": "<SENTINEL_ID_182>",
|
2160 |
+
"single_word": false,
|
2161 |
+
"lstrip": false,
|
2162 |
+
"rstrip": false,
|
2163 |
+
"normalized": false,
|
2164 |
+
"special": true
|
2165 |
+
},
|
2166 |
+
{
|
2167 |
+
"id": 240,
|
2168 |
+
"content": "<SENTINEL_ID_183>",
|
2169 |
+
"single_word": false,
|
2170 |
+
"lstrip": false,
|
2171 |
+
"rstrip": false,
|
2172 |
+
"normalized": false,
|
2173 |
+
"special": true
|
2174 |
+
},
|
2175 |
+
{
|
2176 |
+
"id": 241,
|
2177 |
+
"content": "<SENTINEL_ID_184>",
|
2178 |
+
"single_word": false,
|
2179 |
+
"lstrip": false,
|
2180 |
+
"rstrip": false,
|
2181 |
+
"normalized": false,
|
2182 |
+
"special": true
|
2183 |
+
},
|
2184 |
+
{
|
2185 |
+
"id": 242,
|
2186 |
+
"content": "<SENTINEL_ID_185>",
|
2187 |
+
"single_word": false,
|
2188 |
+
"lstrip": false,
|
2189 |
+
"rstrip": false,
|
2190 |
+
"normalized": false,
|
2191 |
+
"special": true
|
2192 |
+
},
|
2193 |
+
{
|
2194 |
+
"id": 243,
|
2195 |
+
"content": "<SENTINEL_ID_186>",
|
2196 |
+
"single_word": false,
|
2197 |
+
"lstrip": false,
|
2198 |
+
"rstrip": false,
|
2199 |
+
"normalized": false,
|
2200 |
+
"special": true
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"id": 244,
|
2204 |
+
"content": "<SENTINEL_ID_187>",
|
2205 |
+
"single_word": false,
|
2206 |
+
"lstrip": false,
|
2207 |
+
"rstrip": false,
|
2208 |
+
"normalized": false,
|
2209 |
+
"special": true
|
2210 |
+
},
|
2211 |
+
{
|
2212 |
+
"id": 245,
|
2213 |
+
"content": "<SENTINEL_ID_188>",
|
2214 |
+
"single_word": false,
|
2215 |
+
"lstrip": false,
|
2216 |
+
"rstrip": false,
|
2217 |
+
"normalized": false,
|
2218 |
+
"special": true
|
2219 |
+
},
|
2220 |
+
{
|
2221 |
+
"id": 246,
|
2222 |
+
"content": "<SENTINEL_ID_189>",
|
2223 |
+
"single_word": false,
|
2224 |
+
"lstrip": false,
|
2225 |
+
"rstrip": false,
|
2226 |
+
"normalized": false,
|
2227 |
+
"special": true
|
2228 |
+
},
|
2229 |
+
{
|
2230 |
+
"id": 247,
|
2231 |
+
"content": "<SENTINEL_ID_190>",
|
2232 |
+
"single_word": false,
|
2233 |
+
"lstrip": false,
|
2234 |
+
"rstrip": false,
|
2235 |
+
"normalized": false,
|
2236 |
+
"special": true
|
2237 |
+
},
|
2238 |
+
{
|
2239 |
+
"id": 248,
|
2240 |
+
"content": "<SENTINEL_ID_191>",
|
2241 |
+
"single_word": false,
|
2242 |
+
"lstrip": false,
|
2243 |
+
"rstrip": false,
|
2244 |
+
"normalized": false,
|
2245 |
+
"special": true
|
2246 |
+
},
|
2247 |
+
{
|
2248 |
+
"id": 249,
|
2249 |
+
"content": "<SENTINEL_ID_192>",
|
2250 |
+
"single_word": false,
|
2251 |
+
"lstrip": false,
|
2252 |
+
"rstrip": false,
|
2253 |
+
"normalized": false,
|
2254 |
+
"special": true
|
2255 |
+
},
|
2256 |
+
{
|
2257 |
+
"id": 250,
|
2258 |
+
"content": "<SENTINEL_ID_193>",
|
2259 |
+
"single_word": false,
|
2260 |
+
"lstrip": false,
|
2261 |
+
"rstrip": false,
|
2262 |
+
"normalized": false,
|
2263 |
+
"special": true
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"id": 251,
|
2267 |
+
"content": "<SENTINEL_ID_194>",
|
2268 |
+
"single_word": false,
|
2269 |
+
"lstrip": false,
|
2270 |
+
"rstrip": false,
|
2271 |
+
"normalized": false,
|
2272 |
+
"special": true
|
2273 |
+
},
|
2274 |
+
{
|
2275 |
+
"id": 252,
|
2276 |
+
"content": "<SENTINEL_ID_195>",
|
2277 |
+
"single_word": false,
|
2278 |
+
"lstrip": false,
|
2279 |
+
"rstrip": false,
|
2280 |
+
"normalized": false,
|
2281 |
+
"special": true
|
2282 |
+
},
|
2283 |
+
{
|
2284 |
+
"id": 253,
|
2285 |
+
"content": "<SENTINEL_ID_196>",
|
2286 |
+
"single_word": false,
|
2287 |
+
"lstrip": false,
|
2288 |
+
"rstrip": false,
|
2289 |
+
"normalized": false,
|
2290 |
+
"special": true
|
2291 |
+
},
|
2292 |
+
{
|
2293 |
+
"id": 254,
|
2294 |
+
"content": "<SENTINEL_ID_197>",
|
2295 |
+
"single_word": false,
|
2296 |
+
"lstrip": false,
|
2297 |
+
"rstrip": false,
|
2298 |
+
"normalized": false,
|
2299 |
+
"special": true
|
2300 |
+
},
|
2301 |
+
{
|
2302 |
+
"id": 255,
|
2303 |
+
"content": "<SENTINEL_ID_198>",
|
2304 |
+
"single_word": false,
|
2305 |
+
"lstrip": false,
|
2306 |
+
"rstrip": false,
|
2307 |
+
"normalized": false,
|
2308 |
+
"special": true
|
2309 |
+
},
|
2310 |
+
{
|
2311 |
+
"id": 256,
|
2312 |
+
"content": "<SENTINEL_ID_199>",
|
2313 |
+
"single_word": false,
|
2314 |
+
"lstrip": false,
|
2315 |
+
"rstrip": false,
|
2316 |
+
"normalized": false,
|
2317 |
+
"special": true
|
2318 |
+
},
|
2319 |
+
{
|
2320 |
+
"id": 257,
|
2321 |
+
"content": "<MOLECULAR_ENTITY_TYPE_ANTIGEN>",
|
2322 |
+
"single_word": false,
|
2323 |
+
"lstrip": false,
|
2324 |
+
"rstrip": false,
|
2325 |
+
"normalized": false,
|
2326 |
+
"special": true
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"id": 258,
|
2330 |
+
"content": "<MOLECULAR_ENTITY_TYPE_ANTIBODY_LIGHT_CHAIN>",
|
2331 |
+
"single_word": false,
|
2332 |
+
"lstrip": false,
|
2333 |
+
"rstrip": false,
|
2334 |
+
"normalized": false,
|
2335 |
+
"special": true
|
2336 |
+
},
|
2337 |
+
{
|
2338 |
+
"id": 259,
|
2339 |
+
"content": "<MOLECULAR_ENTITY_TYPE_ANTIBODY_HEAVY_CHAIN>",
|
2340 |
+
"single_word": false,
|
2341 |
+
"lstrip": false,
|
2342 |
+
"rstrip": false,
|
2343 |
+
"normalized": false,
|
2344 |
+
"special": true
|
2345 |
+
},
|
2346 |
+
{
|
2347 |
+
"id": 260,
|
2348 |
+
"content": "<ATTRIBUTE_ORGANISM>",
|
2349 |
+
"single_word": false,
|
2350 |
+
"lstrip": false,
|
2351 |
+
"rstrip": false,
|
2352 |
+
"normalized": false,
|
2353 |
+
"special": true
|
2354 |
+
},
|
2355 |
+
{
|
2356 |
+
"id": 261,
|
2357 |
+
"content": "<ATTRIBUTE_ORGANISM_HUMAN>",
|
2358 |
+
"single_word": false,
|
2359 |
+
"lstrip": false,
|
2360 |
+
"rstrip": false,
|
2361 |
+
"normalized": false,
|
2362 |
+
"special": true
|
2363 |
+
},
|
2364 |
+
{
|
2365 |
+
"id": 262,
|
2366 |
+
"content": "<ATTRIBUTE_ORGANISM_RABBIT>",
|
2367 |
+
"single_word": false,
|
2368 |
+
"lstrip": false,
|
2369 |
+
"rstrip": false,
|
2370 |
+
"normalized": false,
|
2371 |
+
"special": true
|
2372 |
+
},
|
2373 |
+
{
|
2374 |
+
"id": 263,
|
2375 |
+
"content": "<ATTRIBUTE_ORGANISM_RAT>",
|
2376 |
+
"single_word": false,
|
2377 |
+
"lstrip": false,
|
2378 |
+
"rstrip": false,
|
2379 |
+
"normalized": false,
|
2380 |
+
"special": true
|
2381 |
+
},
|
2382 |
+
{
|
2383 |
+
"id": 264,
|
2384 |
+
"content": "<ATTRIBUTE_ORGANISM_MOUSE>",
|
2385 |
+
"single_word": false,
|
2386 |
+
"lstrip": false,
|
2387 |
+
"rstrip": false,
|
2388 |
+
"normalized": false,
|
2389 |
+
"special": true
|
2390 |
+
},
|
2391 |
+
{
|
2392 |
+
"id": 265,
|
2393 |
+
"content": "<ATTRIBUTE_ORGANISM_MONKEY>",
|
2394 |
+
"single_word": false,
|
2395 |
+
"lstrip": false,
|
2396 |
+
"rstrip": false,
|
2397 |
+
"normalized": false,
|
2398 |
+
"special": true
|
2399 |
+
},
|
2400 |
+
{
|
2401 |
+
"id": 266,
|
2402 |
+
"content": "<ATTRIBUTE_ORGANISM_CAMEL>",
|
2403 |
+
"single_word": false,
|
2404 |
+
"lstrip": false,
|
2405 |
+
"rstrip": false,
|
2406 |
+
"normalized": false,
|
2407 |
+
"special": true
|
2408 |
+
},
|
2409 |
+
{
|
2410 |
+
"id": 267,
|
2411 |
+
"content": "<EPITOPE_PARATOPE_PREDICTION>",
|
2412 |
+
"single_word": false,
|
2413 |
+
"lstrip": false,
|
2414 |
+
"rstrip": false,
|
2415 |
+
"normalized": false,
|
2416 |
+
"special": true
|
2417 |
+
},
|
2418 |
+
{
|
2419 |
+
"id": 268,
|
2420 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN_CDR1>",
|
2421 |
+
"single_word": false,
|
2422 |
+
"lstrip": false,
|
2423 |
+
"rstrip": false,
|
2424 |
+
"normalized": false,
|
2425 |
+
"special": true
|
2426 |
+
},
|
2427 |
+
{
|
2428 |
+
"id": 269,
|
2429 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN_CDR3>",
|
2430 |
+
"single_word": false,
|
2431 |
+
"lstrip": false,
|
2432 |
+
"rstrip": false,
|
2433 |
+
"normalized": false,
|
2434 |
+
"special": true
|
2435 |
+
},
|
2436 |
+
{
|
2437 |
+
"id": 270,
|
2438 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN_CDR3>",
|
2439 |
+
"single_word": false,
|
2440 |
+
"lstrip": false,
|
2441 |
+
"rstrip": false,
|
2442 |
+
"normalized": false,
|
2443 |
+
"special": true
|
2444 |
+
},
|
2445 |
+
{
|
2446 |
+
"id": 271,
|
2447 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN_CDR2>",
|
2448 |
+
"single_word": false,
|
2449 |
+
"lstrip": false,
|
2450 |
+
"rstrip": false,
|
2451 |
+
"normalized": false,
|
2452 |
+
"special": true
|
2453 |
+
},
|
2454 |
+
{
|
2455 |
+
"id": 272,
|
2456 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN_CDR2>",
|
2457 |
+
"single_word": false,
|
2458 |
+
"lstrip": false,
|
2459 |
+
"rstrip": false,
|
2460 |
+
"normalized": false,
|
2461 |
+
"special": true
|
2462 |
+
},
|
2463 |
+
{
|
2464 |
+
"id": 273,
|
2465 |
+
"content": "<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN_CDR1>",
|
2466 |
+
"single_word": false,
|
2467 |
+
"lstrip": false,
|
2468 |
+
"rstrip": false,
|
2469 |
+
"normalized": false,
|
2470 |
+
"special": true
|
2471 |
+
},
|
2472 |
+
{
|
2473 |
+
"id": 274,
|
2474 |
+
"content": "<MOLECULAR_ENTITY_GENERAL_PROTEIN>",
|
2475 |
+
"single_word": false,
|
2476 |
+
"lstrip": false,
|
2477 |
+
"rstrip": false,
|
2478 |
+
"normalized": false,
|
2479 |
+
"special": true
|
2480 |
+
},
|
2481 |
+
{
|
2482 |
+
"id": 275,
|
2483 |
+
"content": "<TIMESTEP>",
|
2484 |
+
"single_word": false,
|
2485 |
+
"lstrip": false,
|
2486 |
+
"rstrip": false,
|
2487 |
+
"normalized": false,
|
2488 |
+
"special": true
|
2489 |
+
},
|
2490 |
+
{
|
2491 |
+
"id": 276,
|
2492 |
+
"content": "<DIFFUSION>",
|
2493 |
+
"single_word": false,
|
2494 |
+
"lstrip": false,
|
2495 |
+
"rstrip": false,
|
2496 |
+
"normalized": false,
|
2497 |
+
"special": true
|
2498 |
+
},
|
2499 |
+
{
|
2500 |
+
"id": 277,
|
2501 |
+
"content": "<SEQUENCE_NATURAL_END>",
|
2502 |
+
"single_word": false,
|
2503 |
+
"lstrip": false,
|
2504 |
+
"rstrip": false,
|
2505 |
+
"normalized": false,
|
2506 |
+
"special": true
|
2507 |
+
},
|
2508 |
+
{
|
2509 |
+
"id": 278,
|
2510 |
+
"content": "<SMILES_SEQUENCE>",
|
2511 |
+
"single_word": false,
|
2512 |
+
"lstrip": false,
|
2513 |
+
"rstrip": false,
|
2514 |
+
"normalized": false,
|
2515 |
+
"special": true
|
2516 |
+
},
|
2517 |
+
{
|
2518 |
+
"id": 279,
|
2519 |
+
"content": "<SELFIES_SEQUENCE>",
|
2520 |
+
"single_word": false,
|
2521 |
+
"lstrip": false,
|
2522 |
+
"rstrip": false,
|
2523 |
+
"normalized": false,
|
2524 |
+
"special": true
|
2525 |
+
},
|
2526 |
+
{
|
2527 |
+
"id": 280,
|
2528 |
+
"content": "<AMINO_ACID_SEQUENCE>",
|
2529 |
+
"single_word": false,
|
2530 |
+
"lstrip": false,
|
2531 |
+
"rstrip": false,
|
2532 |
+
"normalized": false,
|
2533 |
+
"special": true
|
2534 |
+
},
|
2535 |
+
{
|
2536 |
+
"id": 281,
|
2537 |
+
"content": "<GENERAL_AFFINITY_CLASS>",
|
2538 |
+
"single_word": false,
|
2539 |
+
"lstrip": false,
|
2540 |
+
"rstrip": false,
|
2541 |
+
"normalized": false,
|
2542 |
+
"special": true
|
2543 |
+
},
|
2544 |
+
{
|
2545 |
+
"id": 282,
|
2546 |
+
"content": "<BACKSPACE>",
|
2547 |
+
"single_word": false,
|
2548 |
+
"lstrip": false,
|
2549 |
+
"rstrip": false,
|
2550 |
+
"normalized": false,
|
2551 |
+
"special": true
|
2552 |
+
},
|
2553 |
+
{
|
2554 |
+
"id": 283,
|
2555 |
+
"content": "<SEQUENCE_NATURAL_START>",
|
2556 |
+
"single_word": false,
|
2557 |
+
"lstrip": false,
|
2558 |
+
"rstrip": false,
|
2559 |
+
"normalized": false,
|
2560 |
+
"special": true
|
2561 |
+
},
|
2562 |
+
{
|
2563 |
+
"id": 284,
|
2564 |
+
"content": "<NOOP>",
|
2565 |
+
"single_word": false,
|
2566 |
+
"lstrip": false,
|
2567 |
+
"rstrip": false,
|
2568 |
+
"normalized": false,
|
2569 |
+
"special": true
|
2570 |
+
},
|
2571 |
+
{
|
2572 |
+
"id": 285,
|
2573 |
+
"content": "<TARGETED_ANTIBODY_DESIGN_ENCODER_ONLY_MODE>",
|
2574 |
+
"single_word": false,
|
2575 |
+
"lstrip": false,
|
2576 |
+
"rstrip": false,
|
2577 |
+
"normalized": false,
|
2578 |
+
"special": true
|
2579 |
+
},
|
2580 |
+
{
|
2581 |
+
"id": 286,
|
2582 |
+
"content": "<MOLECULAR_ENTITY_SMALL_MOLECULE>",
|
2583 |
+
"single_word": false,
|
2584 |
+
"lstrip": false,
|
2585 |
+
"rstrip": false,
|
2586 |
+
"normalized": false,
|
2587 |
+
"special": true
|
2588 |
+
},
|
2589 |
+
{
|
2590 |
+
"id": 287,
|
2591 |
+
"content": "<MOLECULAR_ENTITY_CELL_GENE_EXPRESSION_RANKED>",
|
2592 |
+
"single_word": false,
|
2593 |
+
"lstrip": false,
|
2594 |
+
"rstrip": false,
|
2595 |
+
"normalized": false,
|
2596 |
+
"special": true
|
2597 |
+
},
|
2598 |
+
{
|
2599 |
+
"id": 288,
|
2600 |
+
"content": "<CELL_TYPE_CLASS>",
|
2601 |
+
"single_word": false,
|
2602 |
+
"lstrip": false,
|
2603 |
+
"rstrip": false,
|
2604 |
+
"normalized": false,
|
2605 |
+
"special": true
|
2606 |
+
},
|
2607 |
+
{
|
2608 |
+
"id": 289,
|
2609 |
+
"content": "<TISSUE_TYPE_CLASS>",
|
2610 |
+
"single_word": false,
|
2611 |
+
"lstrip": false,
|
2612 |
+
"rstrip": false,
|
2613 |
+
"normalized": false,
|
2614 |
+
"special": true
|
2615 |
+
},
|
2616 |
+
{
|
2617 |
+
"id": 290,
|
2618 |
+
"content": "<CORRUPTED_AREA_START>",
|
2619 |
+
"single_word": false,
|
2620 |
+
"lstrip": false,
|
2621 |
+
"rstrip": false,
|
2622 |
+
"normalized": false,
|
2623 |
+
"special": true
|
2624 |
+
},
|
2625 |
+
{
|
2626 |
+
"id": 291,
|
2627 |
+
"content": "<CORRUPTED_AREA_END>",
|
2628 |
+
"single_word": false,
|
2629 |
+
"lstrip": false,
|
2630 |
+
"rstrip": false,
|
2631 |
+
"normalized": false,
|
2632 |
+
"special": true
|
2633 |
+
},
|
2634 |
+
{
|
2635 |
+
"id": 292,
|
2636 |
+
"content": "<MOLECULAR_ENTITY_MUTATED_PROTEIN_CHAIN>",
|
2637 |
+
"single_word": false,
|
2638 |
+
"lstrip": false,
|
2639 |
+
"rstrip": false,
|
2640 |
+
"normalized": false,
|
2641 |
+
"special": true
|
2642 |
+
},
|
2643 |
+
{
|
2644 |
+
"id": 293,
|
2645 |
+
"content": "<MOLECULAR_ENTITY_PROTEIN_CHAIN>",
|
2646 |
+
"single_word": false,
|
2647 |
+
"lstrip": false,
|
2648 |
+
"rstrip": false,
|
2649 |
+
"normalized": false,
|
2650 |
+
"special": true
|
2651 |
+
},
|
2652 |
+
{
|
2653 |
+
"id": 294,
|
2654 |
+
"content": "<COMPLEX_ENTITY>",
|
2655 |
+
"single_word": false,
|
2656 |
+
"lstrip": false,
|
2657 |
+
"rstrip": false,
|
2658 |
+
"normalized": false,
|
2659 |
+
"special": true
|
2660 |
+
},
|
2661 |
+
{
|
2662 |
+
"id": 295,
|
2663 |
+
"content": "<ALTERNATIVE>",
|
2664 |
+
"single_word": false,
|
2665 |
+
"lstrip": false,
|
2666 |
+
"rstrip": false,
|
2667 |
+
"normalized": false,
|
2668 |
+
"special": true
|
2669 |
+
},
|
2670 |
+
{
|
2671 |
+
"id": 296,
|
2672 |
+
"content": "<CDR3_REGION>",
|
2673 |
+
"single_word": false,
|
2674 |
+
"lstrip": false,
|
2675 |
+
"rstrip": false,
|
2676 |
+
"normalized": false,
|
2677 |
+
"special": true
|
2678 |
+
},
|
2679 |
+
{
|
2680 |
+
"id": 297,
|
2681 |
+
"content": "<GENERAL_CHAIN>",
|
2682 |
+
"single_word": false,
|
2683 |
+
"lstrip": false,
|
2684 |
+
"rstrip": false,
|
2685 |
+
"normalized": false,
|
2686 |
+
"special": true
|
2687 |
+
},
|
2688 |
+
{
|
2689 |
+
"id": 298,
|
2690 |
+
"content": "<SUBMOLECULAR_ENTITY>",
|
2691 |
+
"single_word": false,
|
2692 |
+
"lstrip": false,
|
2693 |
+
"rstrip": false,
|
2694 |
+
"normalized": false,
|
2695 |
+
"special": true
|
2696 |
+
},
|
2697 |
+
{
|
2698 |
+
"id": 299,
|
2699 |
+
"content": "<MUTATED>",
|
2700 |
+
"single_word": false,
|
2701 |
+
"lstrip": false,
|
2702 |
+
"rstrip": false,
|
2703 |
+
"normalized": false,
|
2704 |
+
"special": true
|
2705 |
+
},
|
2706 |
+
{
|
2707 |
+
"id": 300,
|
2708 |
+
"content": "<MOLECULAR_ENTITY_TCR_ALPHA_CDR3>",
|
2709 |
+
"single_word": false,
|
2710 |
+
"lstrip": false,
|
2711 |
+
"rstrip": false,
|
2712 |
+
"normalized": false,
|
2713 |
+
"special": true
|
2714 |
+
},
|
2715 |
+
{
|
2716 |
+
"id": 301,
|
2717 |
+
"content": "<MOLECULAR_ENTITY_TCR_DELTA_CDR3>",
|
2718 |
+
"single_word": false,
|
2719 |
+
"lstrip": false,
|
2720 |
+
"rstrip": false,
|
2721 |
+
"normalized": false,
|
2722 |
+
"special": true
|
2723 |
+
},
|
2724 |
+
{
|
2725 |
+
"id": 302,
|
2726 |
+
"content": "<MOLECULAR_ENTITY_TCR_DELTA_VAR>",
|
2727 |
+
"single_word": false,
|
2728 |
+
"lstrip": false,
|
2729 |
+
"rstrip": false,
|
2730 |
+
"normalized": false,
|
2731 |
+
"special": true
|
2732 |
+
},
|
2733 |
+
{
|
2734 |
+
"id": 303,
|
2735 |
+
"content": "<MOLECULAR_ENTITY_TCR_GAMMA_CDR3>",
|
2736 |
+
"single_word": false,
|
2737 |
+
"lstrip": false,
|
2738 |
+
"rstrip": false,
|
2739 |
+
"normalized": false,
|
2740 |
+
"special": true
|
2741 |
+
},
|
2742 |
+
{
|
2743 |
+
"id": 304,
|
2744 |
+
"content": "<MOLECULAR_ENTITY_TCR_GAMMA_VAR>",
|
2745 |
+
"single_word": false,
|
2746 |
+
"lstrip": false,
|
2747 |
+
"rstrip": false,
|
2748 |
+
"normalized": false,
|
2749 |
+
"special": true
|
2750 |
+
},
|
2751 |
+
{
|
2752 |
+
"id": 305,
|
2753 |
+
"content": "<SCALAR>",
|
2754 |
+
"single_word": false,
|
2755 |
+
"lstrip": false,
|
2756 |
+
"rstrip": false,
|
2757 |
+
"normalized": false,
|
2758 |
+
"special": true
|
2759 |
+
},
|
2760 |
+
{
|
2761 |
+
"id": 306,
|
2762 |
+
"content": "<VECTOR>",
|
2763 |
+
"single_word": false,
|
2764 |
+
"lstrip": false,
|
2765 |
+
"rstrip": false,
|
2766 |
+
"normalized": false,
|
2767 |
+
"special": true
|
2768 |
+
},
|
2769 |
+
{
|
2770 |
+
"id": 307,
|
2771 |
+
"content": "<MASKED_SCALAR>",
|
2772 |
+
"single_word": false,
|
2773 |
+
"lstrip": false,
|
2774 |
+
"rstrip": false,
|
2775 |
+
"normalized": false,
|
2776 |
+
"special": true
|
2777 |
+
},
|
2778 |
+
{
|
2779 |
+
"id": 308,
|
2780 |
+
"content": "<MASKED_VECTOR>",
|
2781 |
+
"single_word": false,
|
2782 |
+
"lstrip": false,
|
2783 |
+
"rstrip": false,
|
2784 |
+
"normalized": false,
|
2785 |
+
"special": true
|
2786 |
+
},
|
2787 |
+
{
|
2788 |
+
"id": 309,
|
2789 |
+
"content": "<AUTOENCODER_LATENT_LOG_VARIANCE>",
|
2790 |
+
"single_word": false,
|
2791 |
+
"lstrip": false,
|
2792 |
+
"rstrip": false,
|
2793 |
+
"normalized": false,
|
2794 |
+
"special": true
|
2795 |
+
},
|
2796 |
+
{
|
2797 |
+
"id": 310,
|
2798 |
+
"content": "<AUTOENCODER_LATENT_MEAN>",
|
2799 |
+
"single_word": false,
|
2800 |
+
"lstrip": false,
|
2801 |
+
"rstrip": false,
|
2802 |
+
"normalized": false,
|
2803 |
+
"special": true
|
2804 |
+
},
|
2805 |
+
{
|
2806 |
+
"id": 311,
|
2807 |
+
"content": "<AUTOENCODER_LATENT_SAMPLED_Z>",
|
2808 |
+
"single_word": false,
|
2809 |
+
"lstrip": false,
|
2810 |
+
"rstrip": false,
|
2811 |
+
"normalized": false,
|
2812 |
+
"special": true
|
2813 |
+
},
|
2814 |
+
{
|
2815 |
+
"id": 312,
|
2816 |
+
"content": "<AUTOENCODER_TASK>",
|
2817 |
+
"single_word": false,
|
2818 |
+
"lstrip": false,
|
2819 |
+
"rstrip": false,
|
2820 |
+
"normalized": false,
|
2821 |
+
"special": true
|
2822 |
+
},
|
2823 |
+
{
|
2824 |
+
"id": 313,
|
2825 |
+
"content": "<DECODED_FROM_LATENT>",
|
2826 |
+
"single_word": false,
|
2827 |
+
"lstrip": false,
|
2828 |
+
"rstrip": false,
|
2829 |
+
"normalized": false,
|
2830 |
+
"special": true
|
2831 |
+
},
|
2832 |
+
{
|
2833 |
+
"id": 314,
|
2834 |
+
"content": "<BBBP>",
|
2835 |
+
"single_word": false,
|
2836 |
+
"lstrip": false,
|
2837 |
+
"rstrip": false,
|
2838 |
+
"normalized": false,
|
2839 |
+
"special": true
|
2840 |
+
},
|
2841 |
+
{
|
2842 |
+
"id": 315,
|
2843 |
+
"content": "<FDA_APPR>",
|
2844 |
+
"single_word": false,
|
2845 |
+
"lstrip": false,
|
2846 |
+
"rstrip": false,
|
2847 |
+
"normalized": false,
|
2848 |
+
"special": true
|
2849 |
+
},
|
2850 |
+
{
|
2851 |
+
"id": 316,
|
2852 |
+
"content": "<HIV_ACTIVITY>",
|
2853 |
+
"single_word": false,
|
2854 |
+
"lstrip": false,
|
2855 |
+
"rstrip": false,
|
2856 |
+
"normalized": false,
|
2857 |
+
"special": true
|
2858 |
+
}
|
2859 |
+
],
|
2860 |
+
"normalizer": null,
|
2861 |
+
"pre_tokenizer": {
|
2862 |
+
"type": "Sequence",
|
2863 |
+
"pretokenizers": [
|
2864 |
+
{
|
2865 |
+
"type": "Split",
|
2866 |
+
"pattern": {
|
2867 |
+
"Regex": "<.*?>|\\S"
|
2868 |
+
},
|
2869 |
+
"behavior": "Removed",
|
2870 |
+
"invert": true
|
2871 |
+
}
|
2872 |
+
]
|
2873 |
+
},
|
2874 |
+
"post_processor": null,
|
2875 |
+
"decoder": null,
|
2876 |
+
"model": {
|
2877 |
+
"type": "WordLevel",
|
2878 |
+
"vocab": {
|
2879 |
+
"<UNK>": 0,
|
2880 |
+
"<PAD>": 1,
|
2881 |
+
"<CLS>": 2,
|
2882 |
+
"<SEP>": 3,
|
2883 |
+
"<MASK>": 4,
|
2884 |
+
"<EOS>": 5,
|
2885 |
+
"<MOLECULAR_ENTITY>": 6,
|
2886 |
+
"<GLOBAL_INTERACTION_ATTRIBUTES>": 7,
|
2887 |
+
"<MOLECULAR_ENTITY_ANTIGEN>": 8,
|
2888 |
+
"<MOLECULAR_ENTITY_EPITOPE>": 9,
|
2889 |
+
"<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN>": 10,
|
2890 |
+
"<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN>": 11,
|
2891 |
+
"<MOLECULAR_ENTITY_TCR_ALPHA_CHAIN>": 12,
|
2892 |
+
"<MOLECULAR_ENTITY_TCR_BETA_VDJ>": 13,
|
2893 |
+
"<MOLECULAR_ENTITY_TCR_BETA_CDR3>": 14,
|
2894 |
+
"<BINDING_AFFINITY_CLASS>": 15,
|
2895 |
+
"<DECODER_START>": 16,
|
2896 |
+
"<BINDING>": 17,
|
2897 |
+
"<FILLIN>": 18,
|
2898 |
+
"<REORDER>": 19,
|
2899 |
+
"<TOAA>": 20,
|
2900 |
+
"<ACTIVE>": 21,
|
2901 |
+
"<GENESEQ>": 22,
|
2902 |
+
"<INCREASE>": 23,
|
2903 |
+
"<DECREASE>": 24,
|
2904 |
+
"<STRUCTURE>": 25,
|
2905 |
+
"<DISTANCE>": 26,
|
2906 |
+
"<SOLUBILITY>": 27,
|
2907 |
+
"<TOXICITY>": 28,
|
2908 |
+
"<AB>": 29,
|
2909 |
+
"<ISACTIVE>": 30,
|
2910 |
+
"<ISSYNTHETIC>": 31,
|
2911 |
+
"<PENETR>": 32,
|
2912 |
+
"<ABSORPTION>": 33,
|
2913 |
+
"<DISTRIBUTION>": 34,
|
2914 |
+
"<METABOLISM>": 35,
|
2915 |
+
"<EXCRETION>": 36,
|
2916 |
+
"<FLUORESCENCE>": 37,
|
2917 |
+
"<STABILITY>": 38,
|
2918 |
+
"<DISORDER>": 39,
|
2919 |
+
"<DISEASE>": 40,
|
2920 |
+
"<BINARY>": 41,
|
2921 |
+
"<REGRESSION>": 42,
|
2922 |
+
"<ORGANISM>": 43,
|
2923 |
+
"<0>": 44,
|
2924 |
+
"<1>": 45,
|
2925 |
+
"<2>": 46,
|
2926 |
+
"<3>": 47,
|
2927 |
+
"<4>": 48,
|
2928 |
+
"<5>": 49,
|
2929 |
+
"<6>": 50,
|
2930 |
+
"<7>": 51,
|
2931 |
+
"<8>": 52,
|
2932 |
+
"<9>": 53,
|
2933 |
+
"<.>": 54,
|
2934 |
+
"<YES>": 55,
|
2935 |
+
"<NO>": 56,
|
2936 |
+
"<SENTINEL_ID_0>": 57,
|
2937 |
+
"<SENTINEL_ID_1>": 58,
|
2938 |
+
"<SENTINEL_ID_2>": 59,
|
2939 |
+
"<SENTINEL_ID_3>": 60,
|
2940 |
+
"<SENTINEL_ID_4>": 61,
|
2941 |
+
"<SENTINEL_ID_5>": 62,
|
2942 |
+
"<SENTINEL_ID_6>": 63,
|
2943 |
+
"<SENTINEL_ID_7>": 64,
|
2944 |
+
"<SENTINEL_ID_8>": 65,
|
2945 |
+
"<SENTINEL_ID_9>": 66,
|
2946 |
+
"<SENTINEL_ID_10>": 67,
|
2947 |
+
"<SENTINEL_ID_11>": 68,
|
2948 |
+
"<SENTINEL_ID_12>": 69,
|
2949 |
+
"<SENTINEL_ID_13>": 70,
|
2950 |
+
"<SENTINEL_ID_14>": 71,
|
2951 |
+
"<SENTINEL_ID_15>": 72,
|
2952 |
+
"<SENTINEL_ID_16>": 73,
|
2953 |
+
"<SENTINEL_ID_17>": 74,
|
2954 |
+
"<SENTINEL_ID_18>": 75,
|
2955 |
+
"<SENTINEL_ID_19>": 76,
|
2956 |
+
"<SENTINEL_ID_20>": 77,
|
2957 |
+
"<SENTINEL_ID_21>": 78,
|
2958 |
+
"<SENTINEL_ID_22>": 79,
|
2959 |
+
"<SENTINEL_ID_23>": 80,
|
2960 |
+
"<SENTINEL_ID_24>": 81,
|
2961 |
+
"<SENTINEL_ID_25>": 82,
|
2962 |
+
"<SENTINEL_ID_26>": 83,
|
2963 |
+
"<SENTINEL_ID_27>": 84,
|
2964 |
+
"<SENTINEL_ID_28>": 85,
|
2965 |
+
"<SENTINEL_ID_29>": 86,
|
2966 |
+
"<SENTINEL_ID_30>": 87,
|
2967 |
+
"<SENTINEL_ID_31>": 88,
|
2968 |
+
"<SENTINEL_ID_32>": 89,
|
2969 |
+
"<SENTINEL_ID_33>": 90,
|
2970 |
+
"<SENTINEL_ID_34>": 91,
|
2971 |
+
"<SENTINEL_ID_35>": 92,
|
2972 |
+
"<SENTINEL_ID_36>": 93,
|
2973 |
+
"<SENTINEL_ID_37>": 94,
|
2974 |
+
"<SENTINEL_ID_38>": 95,
|
2975 |
+
"<SENTINEL_ID_39>": 96,
|
2976 |
+
"<SENTINEL_ID_40>": 97,
|
2977 |
+
"<SENTINEL_ID_41>": 98,
|
2978 |
+
"<SENTINEL_ID_42>": 99,
|
2979 |
+
"<SENTINEL_ID_43>": 100,
|
2980 |
+
"<SENTINEL_ID_44>": 101,
|
2981 |
+
"<SENTINEL_ID_45>": 102,
|
2982 |
+
"<SENTINEL_ID_46>": 103,
|
2983 |
+
"<SENTINEL_ID_47>": 104,
|
2984 |
+
"<SENTINEL_ID_48>": 105,
|
2985 |
+
"<SENTINEL_ID_49>": 106,
|
2986 |
+
"<SENTINEL_ID_50>": 107,
|
2987 |
+
"<SENTINEL_ID_51>": 108,
|
2988 |
+
"<SENTINEL_ID_52>": 109,
|
2989 |
+
"<SENTINEL_ID_53>": 110,
|
2990 |
+
"<SENTINEL_ID_54>": 111,
|
2991 |
+
"<SENTINEL_ID_55>": 112,
|
2992 |
+
"<SENTINEL_ID_56>": 113,
|
2993 |
+
"<SENTINEL_ID_57>": 114,
|
2994 |
+
"<SENTINEL_ID_58>": 115,
|
2995 |
+
"<SENTINEL_ID_59>": 116,
|
2996 |
+
"<SENTINEL_ID_60>": 117,
|
2997 |
+
"<SENTINEL_ID_61>": 118,
|
2998 |
+
"<SENTINEL_ID_62>": 119,
|
2999 |
+
"<SENTINEL_ID_63>": 120,
|
3000 |
+
"<SENTINEL_ID_64>": 121,
|
3001 |
+
"<SENTINEL_ID_65>": 122,
|
3002 |
+
"<SENTINEL_ID_66>": 123,
|
3003 |
+
"<SENTINEL_ID_67>": 124,
|
3004 |
+
"<SENTINEL_ID_68>": 125,
|
3005 |
+
"<SENTINEL_ID_69>": 126,
|
3006 |
+
"<SENTINEL_ID_70>": 127,
|
3007 |
+
"<SENTINEL_ID_71>": 128,
|
3008 |
+
"<SENTINEL_ID_72>": 129,
|
3009 |
+
"<SENTINEL_ID_73>": 130,
|
3010 |
+
"<SENTINEL_ID_74>": 131,
|
3011 |
+
"<SENTINEL_ID_75>": 132,
|
3012 |
+
"<SENTINEL_ID_76>": 133,
|
3013 |
+
"<SENTINEL_ID_77>": 134,
|
3014 |
+
"<SENTINEL_ID_78>": 135,
|
3015 |
+
"<SENTINEL_ID_79>": 136,
|
3016 |
+
"<SENTINEL_ID_80>": 137,
|
3017 |
+
"<SENTINEL_ID_81>": 138,
|
3018 |
+
"<SENTINEL_ID_82>": 139,
|
3019 |
+
"<SENTINEL_ID_83>": 140,
|
3020 |
+
"<SENTINEL_ID_84>": 141,
|
3021 |
+
"<SENTINEL_ID_85>": 142,
|
3022 |
+
"<SENTINEL_ID_86>": 143,
|
3023 |
+
"<SENTINEL_ID_87>": 144,
|
3024 |
+
"<SENTINEL_ID_88>": 145,
|
3025 |
+
"<SENTINEL_ID_89>": 146,
|
3026 |
+
"<SENTINEL_ID_90>": 147,
|
3027 |
+
"<SENTINEL_ID_91>": 148,
|
3028 |
+
"<SENTINEL_ID_92>": 149,
|
3029 |
+
"<SENTINEL_ID_93>": 150,
|
3030 |
+
"<SENTINEL_ID_94>": 151,
|
3031 |
+
"<SENTINEL_ID_95>": 152,
|
3032 |
+
"<SENTINEL_ID_96>": 153,
|
3033 |
+
"<SENTINEL_ID_97>": 154,
|
3034 |
+
"<SENTINEL_ID_98>": 155,
|
3035 |
+
"<SENTINEL_ID_99>": 156,
|
3036 |
+
"<SENTINEL_ID_100>": 157,
|
3037 |
+
"<SENTINEL_ID_101>": 158,
|
3038 |
+
"<SENTINEL_ID_102>": 159,
|
3039 |
+
"<SENTINEL_ID_103>": 160,
|
3040 |
+
"<SENTINEL_ID_104>": 161,
|
3041 |
+
"<SENTINEL_ID_105>": 162,
|
3042 |
+
"<SENTINEL_ID_106>": 163,
|
3043 |
+
"<SENTINEL_ID_107>": 164,
|
3044 |
+
"<SENTINEL_ID_108>": 165,
|
3045 |
+
"<SENTINEL_ID_109>": 166,
|
3046 |
+
"<SENTINEL_ID_110>": 167,
|
3047 |
+
"<SENTINEL_ID_111>": 168,
|
3048 |
+
"<SENTINEL_ID_112>": 169,
|
3049 |
+
"<SENTINEL_ID_113>": 170,
|
3050 |
+
"<SENTINEL_ID_114>": 171,
|
3051 |
+
"<SENTINEL_ID_115>": 172,
|
3052 |
+
"<SENTINEL_ID_116>": 173,
|
3053 |
+
"<SENTINEL_ID_117>": 174,
|
3054 |
+
"<SENTINEL_ID_118>": 175,
|
3055 |
+
"<SENTINEL_ID_119>": 176,
|
3056 |
+
"<SENTINEL_ID_120>": 177,
|
3057 |
+
"<SENTINEL_ID_121>": 178,
|
3058 |
+
"<SENTINEL_ID_122>": 179,
|
3059 |
+
"<SENTINEL_ID_123>": 180,
|
3060 |
+
"<SENTINEL_ID_124>": 181,
|
3061 |
+
"<SENTINEL_ID_125>": 182,
|
3062 |
+
"<SENTINEL_ID_126>": 183,
|
3063 |
+
"<SENTINEL_ID_127>": 184,
|
3064 |
+
"<SENTINEL_ID_128>": 185,
|
3065 |
+
"<SENTINEL_ID_129>": 186,
|
3066 |
+
"<SENTINEL_ID_130>": 187,
|
3067 |
+
"<SENTINEL_ID_131>": 188,
|
3068 |
+
"<SENTINEL_ID_132>": 189,
|
3069 |
+
"<SENTINEL_ID_133>": 190,
|
3070 |
+
"<SENTINEL_ID_134>": 191,
|
3071 |
+
"<SENTINEL_ID_135>": 192,
|
3072 |
+
"<SENTINEL_ID_136>": 193,
|
3073 |
+
"<SENTINEL_ID_137>": 194,
|
3074 |
+
"<SENTINEL_ID_138>": 195,
|
3075 |
+
"<SENTINEL_ID_139>": 196,
|
3076 |
+
"<SENTINEL_ID_140>": 197,
|
3077 |
+
"<SENTINEL_ID_141>": 198,
|
3078 |
+
"<SENTINEL_ID_142>": 199,
|
3079 |
+
"<SENTINEL_ID_143>": 200,
|
3080 |
+
"<SENTINEL_ID_144>": 201,
|
3081 |
+
"<SENTINEL_ID_145>": 202,
|
3082 |
+
"<SENTINEL_ID_146>": 203,
|
3083 |
+
"<SENTINEL_ID_147>": 204,
|
3084 |
+
"<SENTINEL_ID_148>": 205,
|
3085 |
+
"<SENTINEL_ID_149>": 206,
|
3086 |
+
"<SENTINEL_ID_150>": 207,
|
3087 |
+
"<SENTINEL_ID_151>": 208,
|
3088 |
+
"<SENTINEL_ID_152>": 209,
|
3089 |
+
"<SENTINEL_ID_153>": 210,
|
3090 |
+
"<SENTINEL_ID_154>": 211,
|
3091 |
+
"<SENTINEL_ID_155>": 212,
|
3092 |
+
"<SENTINEL_ID_156>": 213,
|
3093 |
+
"<SENTINEL_ID_157>": 214,
|
3094 |
+
"<SENTINEL_ID_158>": 215,
|
3095 |
+
"<SENTINEL_ID_159>": 216,
|
3096 |
+
"<SENTINEL_ID_160>": 217,
|
3097 |
+
"<SENTINEL_ID_161>": 218,
|
3098 |
+
"<SENTINEL_ID_162>": 219,
|
3099 |
+
"<SENTINEL_ID_163>": 220,
|
3100 |
+
"<SENTINEL_ID_164>": 221,
|
3101 |
+
"<SENTINEL_ID_165>": 222,
|
3102 |
+
"<SENTINEL_ID_166>": 223,
|
3103 |
+
"<SENTINEL_ID_167>": 224,
|
3104 |
+
"<SENTINEL_ID_168>": 225,
|
3105 |
+
"<SENTINEL_ID_169>": 226,
|
3106 |
+
"<SENTINEL_ID_170>": 227,
|
3107 |
+
"<SENTINEL_ID_171>": 228,
|
3108 |
+
"<SENTINEL_ID_172>": 229,
|
3109 |
+
"<SENTINEL_ID_173>": 230,
|
3110 |
+
"<SENTINEL_ID_174>": 231,
|
3111 |
+
"<SENTINEL_ID_175>": 232,
|
3112 |
+
"<SENTINEL_ID_176>": 233,
|
3113 |
+
"<SENTINEL_ID_177>": 234,
|
3114 |
+
"<SENTINEL_ID_178>": 235,
|
3115 |
+
"<SENTINEL_ID_179>": 236,
|
3116 |
+
"<SENTINEL_ID_180>": 237,
|
3117 |
+
"<SENTINEL_ID_181>": 238,
|
3118 |
+
"<SENTINEL_ID_182>": 239,
|
3119 |
+
"<SENTINEL_ID_183>": 240,
|
3120 |
+
"<SENTINEL_ID_184>": 241,
|
3121 |
+
"<SENTINEL_ID_185>": 242,
|
3122 |
+
"<SENTINEL_ID_186>": 243,
|
3123 |
+
"<SENTINEL_ID_187>": 244,
|
3124 |
+
"<SENTINEL_ID_188>": 245,
|
3125 |
+
"<SENTINEL_ID_189>": 246,
|
3126 |
+
"<SENTINEL_ID_190>": 247,
|
3127 |
+
"<SENTINEL_ID_191>": 248,
|
3128 |
+
"<SENTINEL_ID_192>": 249,
|
3129 |
+
"<SENTINEL_ID_193>": 250,
|
3130 |
+
"<SENTINEL_ID_194>": 251,
|
3131 |
+
"<SENTINEL_ID_195>": 252,
|
3132 |
+
"<SENTINEL_ID_196>": 253,
|
3133 |
+
"<SENTINEL_ID_197>": 254,
|
3134 |
+
"<SENTINEL_ID_198>": 255,
|
3135 |
+
"<SENTINEL_ID_199>": 256,
|
3136 |
+
"<MOLECULAR_ENTITY_TYPE_ANTIGEN>": 257,
|
3137 |
+
"<MOLECULAR_ENTITY_TYPE_ANTIBODY_LIGHT_CHAIN>": 258,
|
3138 |
+
"<MOLECULAR_ENTITY_TYPE_ANTIBODY_HEAVY_CHAIN>": 259,
|
3139 |
+
"<ATTRIBUTE_ORGANISM>": 260,
|
3140 |
+
"<ATTRIBUTE_ORGANISM_HUMAN>": 261,
|
3141 |
+
"<ATTRIBUTE_ORGANISM_RABBIT>": 262,
|
3142 |
+
"<ATTRIBUTE_ORGANISM_RAT>": 263,
|
3143 |
+
"<ATTRIBUTE_ORGANISM_MOUSE>": 264,
|
3144 |
+
"<ATTRIBUTE_ORGANISM_MONKEY>": 265,
|
3145 |
+
"<ATTRIBUTE_ORGANISM_CAMEL>": 266,
|
3146 |
+
"<EPITOPE_PARATOPE_PREDICTION>": 267,
|
3147 |
+
"<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN_CDR1>": 268,
|
3148 |
+
"<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN_CDR3>": 269,
|
3149 |
+
"<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN_CDR3>": 270,
|
3150 |
+
"<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN_CDR2>": 271,
|
3151 |
+
"<MOLECULAR_ENTITY_ANTIBODY_HEAVY_CHAIN_CDR2>": 272,
|
3152 |
+
"<MOLECULAR_ENTITY_ANTIBODY_LIGHT_CHAIN_CDR1>": 273,
|
3153 |
+
"<MOLECULAR_ENTITY_GENERAL_PROTEIN>": 274,
|
3154 |
+
"<TIMESTEP>": 275,
|
3155 |
+
"<DIFFUSION>": 276,
|
3156 |
+
"<SEQUENCE_NATURAL_END>": 277,
|
3157 |
+
"<SMILES_SEQUENCE>": 278,
|
3158 |
+
"<SELFIES_SEQUENCE>": 279,
|
3159 |
+
"<AMINO_ACID_SEQUENCE>": 280,
|
3160 |
+
"<GENERAL_AFFINITY_CLASS>": 281,
|
3161 |
+
"<BACKSPACE>": 282,
|
3162 |
+
"<SEQUENCE_NATURAL_START>": 283,
|
3163 |
+
"<NOOP>": 284,
|
3164 |
+
"<TARGETED_ANTIBODY_DESIGN_ENCODER_ONLY_MODE>": 285,
|
3165 |
+
"<MOLECULAR_ENTITY_SMALL_MOLECULE>": 286,
|
3166 |
+
"<MOLECULAR_ENTITY_CELL_GENE_EXPRESSION_RANKED>": 287,
|
3167 |
+
"<CELL_TYPE_CLASS>": 288,
|
3168 |
+
"<TISSUE_TYPE_CLASS>": 289,
|
3169 |
+
"<CORRUPTED_AREA_START>": 290,
|
3170 |
+
"<CORRUPTED_AREA_END>": 291,
|
3171 |
+
"<MOLECULAR_ENTITY_MUTATED_PROTEIN_CHAIN>": 292,
|
3172 |
+
"<MOLECULAR_ENTITY_PROTEIN_CHAIN>": 293,
|
3173 |
+
"<COMPLEX_ENTITY>": 294,
|
3174 |
+
"<ALTERNATIVE>": 295,
|
3175 |
+
"<CDR3_REGION>": 296,
|
3176 |
+
"<GENERAL_CHAIN>": 297,
|
3177 |
+
"<SUBMOLECULAR_ENTITY>": 298,
|
3178 |
+
"<MUTATED>": 299,
|
3179 |
+
"<MOLECULAR_ENTITY_TCR_ALPHA_CDR3>": 300,
|
3180 |
+
"<MOLECULAR_ENTITY_TCR_DELTA_CDR3>": 301,
|
3181 |
+
"<MOLECULAR_ENTITY_TCR_DELTA_VAR>": 302,
|
3182 |
+
"<MOLECULAR_ENTITY_TCR_GAMMA_CDR3>": 303,
|
3183 |
+
"<MOLECULAR_ENTITY_TCR_GAMMA_VAR>": 304,
|
3184 |
+
"<SCALAR>": 305,
|
3185 |
+
"<VECTOR>": 306,
|
3186 |
+
"<MASKED_SCALAR>": 307,
|
3187 |
+
"<MASKED_VECTOR>": 308,
|
3188 |
+
"<AUTOENCODER_LATENT_LOG_VARIANCE>": 309,
|
3189 |
+
"<AUTOENCODER_LATENT_MEAN>": 310,
|
3190 |
+
"<AUTOENCODER_LATENT_SAMPLED_Z>": 311,
|
3191 |
+
"<AUTOENCODER_TASK>": 312,
|
3192 |
+
"<DECODED_FROM_LATENT>": 313,
|
3193 |
+
"<BBBP>": 314,
|
3194 |
+
"<FDA_APPR>": 315,
|
3195 |
+
"<HIV_ACTIVITY>": 316,
|
3196 |
+
"A": 501,
|
3197 |
+
"B": 502,
|
3198 |
+
"C": 503,
|
3199 |
+
"D": 504,
|
3200 |
+
"E": 505,
|
3201 |
+
"F": 506,
|
3202 |
+
"G": 507,
|
3203 |
+
"H": 508,
|
3204 |
+
"I": 509,
|
3205 |
+
"K": 510,
|
3206 |
+
"L": 511,
|
3207 |
+
"M": 512,
|
3208 |
+
"N": 513,
|
3209 |
+
"O": 514,
|
3210 |
+
"P": 515,
|
3211 |
+
"Q": 516,
|
3212 |
+
"R": 517,
|
3213 |
+
"S": 518,
|
3214 |
+
"T": 519,
|
3215 |
+
"U": 520,
|
3216 |
+
"V": 521,
|
3217 |
+
"W": 522,
|
3218 |
+
"X": 523,
|
3219 |
+
"Y": 524,
|
3220 |
+
"Z": 525,
|
3221 |
+
":": 526
|
3222 |
+
},
|
3223 |
+
"unk_token": "<UNK>"
|
3224 |
+
}
|
3225 |
+
}
|