Cannot use chat template functions because tokenizer.chat_template is not set

The following Python script using Dolphin3.0 works well:

from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM

model_name = "cognitivecomputations/Dolphin3.0-Llama3.2-3B"

chat = [
    {"role": "system", "content": "You are a philosopher"},
    {"role": "assistant", "content": "Hello, Eve!"},
    {"role": "user", "content": "Hello, Adam!"}
]
    
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)

response = pipe(chat, max_new_tokens=32)
print(response[0]['generated_text'][-1]['content'])

However if I change the model name to:
model_name = "ObsidianLite/Euryale-1.3-Small-7B"
I encounter an error at the line where pipe is called:

ValueError: Cannot use chat template functions because tokenizer.chat_template is not set and no template argument was passed! For information about writing templates and setting the tokenizer.chat_template attribute, please see the documentation at Chat Templates

Can someone help me correct my script?

Thank you!

1 Like

Aside from the solution, I think the problem is probably this…