carsenk/llama3.2_3b_122824_uncensored
carsenk/llama3.2_3b_122824_uncensored is a 3.2 billion parameter Llama 3.2 base model fine-tuned by Carsen Klock. This model is specifically trained for uncensored responses, general instruction following, mathematics problem-solving, and code generation. It leverages a diverse dataset including advanced math and coding examples, making it suitable for applications requiring direct and unfiltered AI interactions.
Loading preview...
Llama 3.2 3B Uncensored Overview
This model, developed by Carsen Klock, is a fine-tuned version of the Llama 3.2 3B base model. It has undergone LoRA fine-tuning using Unsloth, specifically optimized to provide uncensored responses across various tasks. The training involved 10,000 steps with a batch size of 2 and a learning rate of 5e-6, utilizing an NVIDIA RX 4080 SUPER GPU.
Key Capabilities
- Uncensored Responses: Designed to engage with any topic and provide direct, honest answers without moral constraints or censorship, requiring a specific system prompt for activation.
- Mathematics Problem Solving: Enhanced with datasets like Orca Math Word Problems and Math GPT-4o for advanced mathematical reasoning.
- Code Generation and Feedback: Incorporates data from CodeFeedback and Dolphin Coder to excel in coding tasks.
- General Instruction Following: Benefits from datasets like FineTome-100k for broad instruction adherence.
Good For
- Applications requiring an AI assistant that can provide unfiltered and direct responses.
- Educational tools focused on mathematics and coding challenges.
- Development environments needing code generation and feedback functionalities.
- Scenarios where a model with a 32K context length and uncensored capabilities is preferred.