How should I set the tokenizer parameter?

[hanquanjushi]

I'm using a remote server and I'm encountering an error when using the "Qwen/Qwen2.5-32B-Instruct" tokenizer with my LMQL model. The error message is as follows:

lmql.runtime.tokenizer.TokenizerNotAvailableError: Failed to locate a suitable tokenizer implementation for 'Qwen/Qwen2.5-32B-Instruct' (Make sure your current environment provides a tokenizer backend like 'transformers', 'tiktoken' or 'llama.cpp' for this model)

Below is the code I'm using:

import lmql

model = lmql.model(
"openai/qwen2.5-32b",
endpoint="http://XXX.XX.XX.XX:XXXXX/v1/chat/completions",
api_key="test",
tokenizer="Qwen/Qwen2.5-32B-Instruct"
)

@lmql.query(model=model)
def tell_a_joke():
'''lmql
"""A great good dad joke. A indicates the punchline
Q:[JOKE]
A:[PUNCHLINE]""" where STOPS_AT(JOKE, "?") and
STOPS_AT(PUNCHLINE, "\n")
'''

tell_a_joke()

Could someone please advise on the correct way to configure the tokenizer parameter? Should I be using a different tokenizer string or is there a specific tokenizer backend that I need to install/configure in my environment?

Any help or guidance would be greatly appreciated!