Is added_tokens.json missing?
#7
by
Arvor
- opened
When converting this model to gguf via llama.cpp, I get a vocab mismatch due to a missing added_tokens.json.
It works fine if I add a "added_tokens.json" with:
{
"<|im_end|>": 32000,
"<|im_start|>": 32001
}