[2025.06.10]train en tokenizer

This commit is contained in:
root
2025-06-10 17:36:53 +08:00
parent c6e171e8ed
commit eb1c6c0c17
13 changed files with 243835 additions and 57395 deletions
@@ -1,8 +1,19 @@
{
"additional_special_tokens": [
"<|im_start|>",
"<|im_end|>"
"<|im_end|>",
"<|object_ref_start|>",
"<|object_ref_end|>",
"<|box_start|>",
"<|box_end|>",
"<|quad_start|>",
"<|quad_end|>",
"<|vision_start|>",
"<|vision_end|>",
"<|vision_pad|>",
"<|image_pad|>",
"<|video_pad|>"
],
"eos_token": "<|endoftext|>",
"eos_token": "<|im_end|>",
"pad_token": "<|endoftext|>"
}