AIbase
EN
Home
Denker Mistral Nemo 12B
D

Denker Mistral Nemo 12B

Developed by nbeerbower
Denker is a small, uncensored, reasoning-focused model, fine-tuned from mistral-nemo-kartoffel-12B using ORPO and QLoRA.
Large Language Model
Transformers
Open Source License:Apache-2.0#Uncensored reasoning#Chain-of-thought optimization#Multi-domain DPO fine-tuning
Downloads 20
Release Time : 5/12/2025

Model Overview

This is an experimental model focused on reasoning and chain-of-thought capabilities, designed without censorship, suitable for tasks requiring logical reasoning and problem-solving.

Model Features

Uncensored design
The model is designed without censorship, suitable for scenarios requiring free expression and exploratory thinking
Reasoning optimization
Focused on logical reasoning and problem-solving abilities, with special emphasis on chain-of-thought performance
ORPO fine-tuning
Fine-tuned using the ORPO (Odds Ratio Preference Optimization) method
QLoRA efficient training
Utilizes QLoRA technology for parameter-efficient fine-tuning

Model Capabilities

Text generation
Logical reasoning
Mathematical problem solving
Code reasoning
Physics reasoning
Chain-of-thought reasoning

Use Cases

Education and Research
Mathematical problem solving
Solving complex mathematical reasoning problems
Physics reasoning
Understanding and solving physics-related problems
Programming assistance
Code reasoning
Helping understand and generate code
General reasoning
Logical problem solving
Solving problems requiring complex logical reasoning
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase