The Phi-3-Mini-4K-Instruct is a 3.8B parameters, lightweight, state-of-the-art open model trained with the Phi-3 datasets that includes both synthetic data and the filtered publicly available websites data with a focus on high-quality and reasoning dense properties.

Uses

Direct Use

python

from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

tokenizer = AutoTokenizer.from_pretrained("Aryan-401/Aryan-401/phi-3-mini-4k-instruct-finetune-guanaco-PEFT-Merged", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("Aryan-401/Aryan-401/phi-3-mini-4k-instruct-finetune-guanaco-PEFT-Merged", trust_remote_code=True)

# Prompt content: "hi"
messages = [
    {"role": "user", "content": "What is the Value of Pi?"}
]
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

model = model.to(device).eval()

input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt')
output_ids = model.generate(input_ids.to(device), max_length= 1000)
response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True)

print(response)
# Pi is an irrational number, which means it cannot be expressed as a simple fraction and its decimal representation goes on forever without repeating. However, the value of Pi is approximately 3.14159. It is often rounded to 3.14 for simplicity in calculations.