Granite 8b Qiskit

Table of Contents

Model Overview

Meet the granite-8b-qiskit model, a powerful tool for generating high-quality quantum computing code using Qiskit. This model is designed to assist both quantum computing practitioners and new Qiskit users in building Qiskit code or responding to Qiskit coding-related instructions and questions.

Tasks and rank

Text-to-Code Generation 630th

Code Generation 788th

Code Completion 761st

Learn More About the Ranking System

Capabilities

The granite-8b-qiskit model excels at generating quantum computing code using Qiskit, assisting with building Qiskit code, and responding to Qiskit-related questions and instructions. With its advanced architecture and massive parameter count (8B parameters), this model can process and generate code at an impressive speed.

Primary Tasks

• Generating quantum computing code using Qiskit
• Assisting with building Qiskit code
• Responding to Qiskit-related questions and instructions

Strengths

This model has several strengths that set it apart from other AI models:

• Improved capabilities: The granite-8b-qiskit model is fine-tuned on top of a massive dataset of Qiskit code and instruction data to improve its capabilities.
• High-quality code generation: The model is designed to generate high-quality, non-deprecated Qiskit code.
• Versatility: The model can be used by both quantum computing professionals and new Qiskit users.

Alternatives

DeepSeek Coder V2 Instruct 442 Code intelligence model

OpenCodeInterpreter CL 70B 24 Code generation model

Performance

So, how does granite-8b-qiskit perform in terms of speed, accuracy, and efficiency? Let’s take a closer look.

Speed

granite-8b-qiskit is trained on a massive dataset of Qiskit code and instruction data, which enables it to generate code at an incredible pace. But what does that mean in practice? For example, can it generate a random circuit with 5 qubits in a matter of seconds? The answer is yes!

Accuracy

Accuracy is crucial when it comes to generating code. A single mistake can lead to errors and bugs that can be difficult to track down. granite-8b-qiskit has been fine-tuned to improve its accuracy in generating high-quality Qiskit code.

Efficiency

Efficiency is another key aspect of granite-8b-qiskit’s performance. The model is designed to be efficient in terms of computational resources, making it accessible to a wide range of users.

Limitations

While granite-8b-qiskit is a powerful tool, it’s not perfect. Let’s take a closer look at some of its limitations.

Lack of Safety Alignment

Unlike some other models, granite-8b-qiskit hasn’t undergone safety alignment training. This means it may produce problematic outputs, especially in critical situations.

Hallucination Risk

Smaller models like granite-8b-qiskit might be more susceptible to hallucination in generation scenarios.

Malicious Utilization

Like all Large Language Models, granite-8b-qiskit can be used for malicious purposes.

Examples

Create a quantum circuit with 3 qubits, apply Hadamard gate to the first qubit, and measure all qubits. from qiskit import QuantumCircuit qc = QuantumCircuit(3) qc.h(0) qc.measure([0, 1, 2], [0, 1, 2])

Generate a Qiskit code snippet to apply a Pauli X gate to the second qubit of a 4-qubit quantum register. from qiskit import QuantumCircuit qr = QuantumCircuit(4) qr.x(1)

Write a Qiskit function to create a quantum circuit that adds two qubits. from qiskit import QuantumCircuit def add_qubits(): qc = QuantumCircuit(2) qc.cx(0, 1) return qc

Example Use Case

Here’s an example of how to use granite-8b-qiskit to generate Qiskit code:

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

# Load the pre-trained model and tokenizer
model_path = "qiskit/granite-8b-qiskit"
tokenizer = AutoTokenizer.from_pretrained(model_path)
model = AutoModelForCausalLM.from_pretrained(model_path)

# Define the input text
input_text = "Build a random circuit with 5 qubits"

# Tokenize the input text
input_tokens = tokenizer(input_text, return_tensors="pt")

# Generate output tokens
output = model.generate(**input_tokens, max_new_tokens=128)

# Decode output tokens into text
output = tokenizer.batch_decode(output)

# Print the output
print(output)

This example demonstrates how to use granite-8b-qiskit to generate Qiskit code based on user input. Simply change the input text to your desired prompt, and the model will generate the corresponding code.