from peft import PeftModel
from transformers import AutoModelForCausalLM
from peft import AutoPeftModelForCausalLM
from transformers import AutoTokenizer
import torch

peft_model_id = "Abdulvajid/Llama_Qlora_Reasoning_ToolCalling_Finetuned_4Bit"

model = AutoPeftModelForCausalLM.from_pretrained(peft_model_id,
                                                 torch_dtype=torch.float16,
                                                 device_map='cuda',
                                                 load_in_4bit=True)

tokenizer = AutoTokenizer.from_pretrained(peft_model_id)

tools = [
    {
        "type": "function",
        "function": {
            "name": "search_restaurants",
            "description": "Search for restaurants in a specific location",
            "parameters": {
                "type": "object",
                "properties": {
                    "location": {
                        "type": "string",
                        "description": "The location to search for restaurants"
                    },
                    "cuisine": {
                        "type": "string",
                        "description": "The cuisine type to filter the restaurants"
                    },
                    "price_range": {
                        "type": "integer",
                        "description": "The price range of the restaurants (1 = cheap to 4 = very expensive)"
                    }
                },
                "required": ["location"]
            }
        }
    }
]

messages=[
        {"role": "user", "content": "I'm in Malappuram, can you find a restaurant for me?"}
    ]

prompt = tokenizer.apply_chat_template(
    messages,
    tools=tools,
    add_generation_prompt=True,
    tokenize=True,
    return_tensors="pt"
).to('cuda')

output = model.generate(prompt, max_new_tokens=500)

print(''.join(tokenizer.batch_decode(output[0][len(prompt[0]):])))

Training procedure

This model was trained with SFT.

Framework versions

TRL: 0.19.1
Transformers: 4.53.1
Pytorch: 2.6.0+cu124
Datasets: 4.0.0
Tokenizers: 0.21.2

Citations

Cite TRL as:

bibtex

@misc{vonwerra2022trl,
	title        = {{TRL: Transformer Reinforcement Learning}},
	author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallou{\'e}dec},
	year         = 2020,
	journal      = {GitHub repository},
	publisher    = {GitHub},
	howpublished = {\url{https://github.com/huggingface/trl}}
}