fix: meta reference + llama4 tokenizer fix

This commit is contained in:
Ashwin Bharambe 2025-04-09 00:46:02 -07:00
parent 10882bf478
commit 8001c30a4f
2 changed files with 5 additions and 3 deletions

View file

@ -56,9 +56,11 @@ LLAMA4_TEXT_POST_TRAIN_SPECIAL_TOKENS = [
"<|text_post_train_reserved_special_token_3|>",
"<|text_post_train_reserved_special_token_4|>",
"<|text_post_train_reserved_special_token_5|>",
"<|text_post_train_reserved_special_token_6|>",
"<|text_post_train_reserved_special_token_7|>",
"<|finetune_right_pad|>",
] + get_reserved_special_tokens(
"text_post_train", 61, 6
"text_post_train", 61, 8
) # <|text_post_train_reserved_special_token_6|>, ..., <|text_post_train_reserved_special_token_66|>
# 200080, ..., 201133