teknium

/

OpenHermes-2.5-Mistral-7B

like 799

Example code:

model_name = 'teknium/OpenHermes-2.5-Mistral-7B'
messages = [
    {"role": "user", "content": "What is your favourite condiment?"},
    {"role": "assistant", "content": "Well, I’m quite partial to a good squeeze of fresh lemon juice. It adds just the right amount of zesty flavour to whatever I’m   cooking up in the kitchen!"},
    {"role": "user", "content": "Do you have mayonnaise recipes?"}
]
tokenizer = LlamaTokenizer.from_pretrained(model_name, trust_remote_code=True, use_fast=True)
encoded = tokenizer.apply_chat_template(messages, return_tensors="pt", add_generation_prompt=True)
model_inputs = encoded.to(device)
generated_ids = model.generate(model_inputs, max_new_tokens=250, do_sample=False, eos_token_id=tokenizer.eos_token_id, pad_token_id = tokenizer.pad_token_id)
decoded = tokenizer.batch_decode(generated_ids, skip_special_tokens=True, clean_up_tokenization_space=True)
print(decoded[0])

Response:

<|im_start|> user
What is your favourite condiment? 
 <|im_start|> assistant
Well, I’m quite partial to a good squeeze of fresh lemon juice. It adds just the right amount of zesty flavour to whatever I’m cooking up in the kitchen! 
 <|im_start|> user
Do you have mayonnaise recipes? 
 <|im_start|> assistant
Of course! Here are two simple recipes for homemade mayonnaise that you can try out. [...]

Is there a way to prevent or suppress the prompt from appearing in the response? Thank you.