summaryrefslogtreecommitdiff
path: root/tokenizer/bpe/special.go
diff options
context:
space:
mode:
authorJonas Knobloch <jonas.knobloch@t-online.de>2026-09-11 18:38:02 +0200
committerJonas Knobloch <jonas.knobloch@t-online.de>2026-09-11 18:38:02 +0200
commit330c6387962b197c5ca7a8051d4864feaecc0f25 (patch)
treef13c82306ab432de6a610b1006d7cc10373dc7f6 /tokenizer/bpe/special.go
parent6e6a94b12d6d43c1ce4b8e587742627af6d86a43 (diff)
WIP
Diffstat (limited to 'tokenizer/bpe/special.go')
-rw-r--r--tokenizer/bpe/special.go18
1 files changed, 18 insertions, 0 deletions
diff --git a/tokenizer/bpe/special.go b/tokenizer/bpe/special.go
new file mode 100644
index 0000000..405199f
--- /dev/null
+++ b/tokenizer/bpe/special.go
@@ -0,0 +1,18 @@
+package bpe
+
+// Add helper to ensure special token
+// wip-special-tokens
+
+func EnsureSpecialToken(t *Tokenizer, token string) int {
+ vocab := Vocab(t)
+
+ for i, v := range vocab {
+ if v == token {
+ return i
+ }
+ }
+
+ AddToken(t, token)
+
+ return len(vocab) - 1
+}