README.md

Llama-3.2-3B-Instruct-function-calling-gorilla-style-5epochs

Model Description

This is a fine-tuned version of unsloth/Llama-3.2-3B-Instruct for function calling capabilities, trained in a Gorilla-style format on a custom dataset. The model is trained to understand function calling instructions and generate appropriate responses.

Training Parameters

• Base Model: unsloth/Llama-3.2-3B-Instruct
• Dataset: Custom Function Calling Dataset (Gorilla-style format)
• Training Type: Supervised Fine-tuning with LoRA
• Epochs: 5

Dataset Format

The model was trained on a custom dataset with the following structure:

{
    "Instruction": "User instruction/query",
    "Functions": ["Available function definitions"],
    "Output": ["Model's response with function calls"]
}

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("BluebrainAI/Llama-3.2-3B-Instruct-function-calling-gorilla-style-5epochs")
tokenizer = AutoTokenizer.from_pretrained("BluebrainAI/Llama-3.2-3B-Instruct-function-calling-gorilla-style-5epochs")

# Example usage
instruction = "Your instruction here"
chat = [
    {"role": "user", "content": instruction}
]
input_text = tokenizer.apply_chat_template(chat, tokenize=False, add_generation_prompt=False)
inputs = tokenizer(input_text, return_tensors="pt", truncation=True)
outputs = model.generate(**inputs, max_new_tokens=512, temperature=0.7)
response = tokenizer.decode(outputs[0], skip_special_tokens=True)

License

This model inherits the license of the base model unsloth/Llama-3.2-3B-Instruct.