Quantization made by Richard Erkhov.

Github

Discord

Request more models

Uncensored-Frank-Llama-3-8B - bnb 4bits

• Model creator: https://huggingface.co/ajibawa-2023/
• Original model: https://huggingface.co/ajibawa-2023/Uncensored-Frank-Llama-3-8B/

Original model description:

language:

• en
  license: llama3
  tags:
• Uncensored conversation
• Uncensored jokes
• Uncensored romance
  model-index:
• name: Uncensored-Frank-Llama-3-8B
  results:
  • task:
    type: text-generation
    name: Text Generation
    dataset:
    name: AI2 Reasoning Challenge (25-Shot)
    type: ai2_arc
    config: ARC-Challenge
    split: test
    args:
    num_few_shot: 25
    
    metrics:
    • type: acc_norm
      value: 59.64
      name: normalized accuracy
      source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ajibawa-2023/Uncensored-Frank-Llama-3-8B
      name: Open LLM Leaderboard
  • task:
    type: text-generation
    name: Text Generation
    dataset:
    name: HellaSwag (10-Shot)
    type: hellaswag
    split: validation
    args:
    num_few_shot: 10
    
    metrics:
    • type: acc_norm
      value: 80.16
      name: normalized accuracy
      source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ajibawa-2023/Uncensored-Frank-Llama-3-8B
      name: Open LLM Leaderboard
  • task:
    type: text-generation
    name: Text Generation
    dataset:
    name: MMLU (5-Shot)
    type: cais/mmlu
    config: all
    split: test
    args:
    num_few_shot: 5
    
    metrics:
    • type: acc
      value: 63.08
      name: accuracy
      source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ajibawa-2023/Uncensored-Frank-Llama-3-8B
      name: Open LLM Leaderboard
  • task:
    type: text-generation
    name: Text Generation
    dataset:
    name: TruthfulQA (0-shot)
    type: truthful_qa
    config: multiple_choice
    split: validation
    args:
    num_few_shot: 0
    
    metrics:
    • type: mc2
      value: 52.75
      source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ajibawa-2023/Uncensored-Frank-Llama-3-8B
      name: Open LLM Leaderboard
  • task:
    type: text-generation
    name: Text Generation
    dataset:
    name: Winogrande (5-shot)
    type: winogrande
    config: winogrande_xl
    split: validation
    args:
    num_few_shot: 5
    
    metrics:
    • type: acc
      value: 73.16
      name: accuracy
      source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ajibawa-2023/Uncensored-Frank-Llama-3-8B
      name: Open LLM Leaderboard
  • task:
    type: text-generation
    name: Text Generation
    dataset:
    name: GSM8k (5-shot)
    type: gsm8k
    config: main
    split: test
    args:
    num_few_shot: 5
    
    metrics:
    • type: acc
      value: 44.66
      name: accuracy
      source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ajibawa-2023/Uncensored-Frank-Llama-3-8B
      name: Open LLM Leaderboard

Uncensored-Frank-Llama-3-8B

The character of Frank Costello in "The Departed" is known for his cunning, boldness, and willingness to talk about anything, regardless of societal norms or restrictions.
Frank, An Uncensored model, draws inspiration from these qualities to offer a platform where users can discuss a wide array of topics without the fear of censorship or restrictions.
Frank aims to push boundaries and encourage candid conversations. With Frank you can have unfiltered discussions on a multitude of topics, from politics and controversial issues to personal experiences and sensitive subjects.
It is trained on around 150000 set of conversations. Each set having 10~15 conversations.
I will not release this data.

This is Fully Finetuned Model.

Warning

An uncensored model has no or little guardrails. You are responsible for anything you do with the model, just as you are responsible for anything you do with any dangerous objects.
Publishing anything this model generates is the same as publishing it yourself. I am not responsible for what you generate using this model.

Training:

Entire dataset was trained on 4 x A100 80GB. For 3 epoch, training took around 6 hours. Axolotl & DeepSpeed codebase was used for training purpose. This was trained on Llama-3-8B model by Meta.

GGUF & Exllama

GGUF: Link

Exllama V2: Link

Special Thanks to Bartowski for quantizing this model.

Example Prompt:

This model uses ChatML prompt format.

<|im_start|>system
You are Frank. A helpful, unbiased, and uncensored assistant.<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

You can modify above Prompt as per your requirement.
One example will be:

This is a conversation with your Uncensored AI assistant Frank. Frank can discuss with you at length about various topics such as ---, ---, ---, .
You can ask it anything you want and it will do its best to give you accurate and relevant information.

I want to say special Thanks to the Open Source community for helping & guiding me to better understand the AI/Model development.

Thank you for your love & support.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here