Free and Open-Source Notbad v1.0 Mistral 24B Model - Empowering Mathematical and Python Programming Reasoning

Notbad V1 0 Mistral 24b

Developed by notbadai

Notbad v1.0 Mistral 24B is a model focused on mathematical and Python programming reasoning, based on Mistral-Small-24B-Instruct-2501 and further trained with reinforcement learning.

Large Language Model

Transformers

Open Source License:Apache-2.0 #Mathematical Reasoning Optimization #Python Programming Enhancement #Concise Reasoning Output

Downloads 29

Release Time : 3/29/2025

Model Overview

This model specializes in reasoning capabilities in the fields of mathematics and programming, capable of generating more concise reasoning outputs and improving performance through self-improvement and reinforcement learning techniques.

Model Features

Mathematical and Programming Reasoning

Specializes in reasoning capabilities in mathematics and Python programming, capable of handling complex mathematical problems and programming tasks.

Concise Reasoning Output

The model can generate more concise reasoning outputs, improving readability and efficiency.

Reinforcement Learning Training

Trained using reinforcement learning techniques, enhancing model performance through self-improvement.

Model Capabilities

Mathematical Reasoning

Programming Code Generation

Text Generation

Instruction Following

Use Cases

Education and Research

Mathematical Problem Solving

Solves complex mathematical problems, providing detailed reasoning steps.

Scored 0.752 on the math benchmark, outperforming several similar models.

Programming Assistance

Generates Python code to help developers quickly implement functionalities.

Scored 0.869 on the humaneval benchmark, demonstrating excellent performance.

🚀 Model Card for Notbad v1.0 Mistral 24B

Notbad v1.0 Mistral 24B is a reasoning model trained in math and Python coding. It builds on the Mistral-Small-24B-Instruct-2501 base model and further enhances its performance through reinforcement learning on math and coding tasks.

✨ Features

Shorter and Cleaner Reasoning Outputs: One of the key features of Notbad v1.0 is its ability to generate concise and clear reasoning outputs.
Self - Improvement Reasoning: The reasoning capabilities of this model come from self - improvement, not from distilling other models. It is fine - tuned from data sampled from multiple RL models starting with the Mistral - Small - 24B - Instruct - 2501.
Use of Open Datasets and RL Techniques: It uses open datasets and reinforcement learning techniques similar to those in Quiet Star and Dr. GRPO.

📚 Documentation

New Model Announcement

📣 New model available: Notbad v1.1 Mistral 24B

Acknowledgments

Special thanks to Lambda and Deep Infra for providing help with compute resources for our research and training this model.

Try the Model

You can try the model on chat.labml.ai.

📊 Benchmark results

General Evaluation

Evaluation	notbad_v1_0_mistral_24b	mistral-small-24B-instruct-2501	gemma-2b-27b	llama-3.3-70b	qwen2.5-32b	gpt-4o-mini-2024-07-18
mmlu_pro	0.642	0.663	0.536	0.666	0.683	0.617
gpqa_main	0.447	0.453	0.344	0.531	0.404	0.377

Math & Coding

Evaluation	notbad_v1_0_mistral_24b	mistral-small-24B-instruct-2501	gemma-2b-27b	llama-3.3-70b	qwen2.5-32b	gpt-4o-mini-2024-07-18
humaneval	0.869	0.848	0.732	0.854	0.909	0.890
math	0.752	0.706	0.535	0.743	0.819	0.761

Instruction following

Evaluation	notbad_v1_0_mistral_24b	mistral-small-24B-instruct-2501	gemma-2b-27b	llama-3.3-70b	qwen2.5-32b	gpt-4o-mini-2024-07-18
ifeval	0.514	0.829	0.8065	0.8835	0.8401	0.8499

Note:

Benchmarks are from Mistral-Small-24B-Instruct-2501 Model Card

📄 License

The model is licensed under Apache-2.0.

Property	Details
Model Type	Reasoning model for math and Python coding
Base Model	Mistral-Small-24B-Instruct-2501
Pipeline Tag	text - generation
Library Name	transformers

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご