metadata

base_model:
  - Orenguteng/Llama-3.1-8B-Lexi-Uncensored-V2
  - arcee-ai/Llama-3.1-SuperNova-Lite
  - VAGOsolutions/Llama-3.1-SauerkrautLM-8b-Instruct
  - unsloth/Llama-3.1-Storm-8B
  - DavidAU/L3.1-Dark-Planet-SpinFire-Uncensored-8B
  - unsloth/Meta-Llama-3.1-8B-Instruct
library_name: transformers
tags:
  - mergekit
  - merge
  - llama
  - llama3.1
  - function calling
  - roleplay
  - uncensored
license: llama3.1
model-index:
  - name: ZEUS-8B-V17-abliterated-V4
    results:
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: IFEval (0-Shot)
          type: wis-k/instruction-following-eval
          split: train
          args:
            num_few_shot: 0
        metrics:
          - type: inst_level_strict_acc and prompt_level_strict_acc
            value: 72.28
            name: averaged accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=T145%2FZEUS-8B-V17-abliterated-V4
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: BBH (3-Shot)
          type: SaylorTwift/bbh
          split: test
          args:
            num_few_shot: 3
        metrics:
          - type: acc_norm
            value: 30.97
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=T145%2FZEUS-8B-V17-abliterated-V4
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MATH Lvl 5 (4-Shot)
          type: lighteval/MATH-Hard
          split: test
          args:
            num_few_shot: 4
        metrics:
          - type: exact_match
            value: 1.81
            name: exact match
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=T145%2FZEUS-8B-V17-abliterated-V4
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: GPQA (0-shot)
          type: Idavidrein/gpqa
          split: train
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 4.47
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=T145%2FZEUS-8B-V17-abliterated-V4
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MuSR (0-shot)
          type: TAUR-Lab/MuSR
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 11.61
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=T145%2FZEUS-8B-V17-abliterated-V4
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MMLU-PRO (5-shot)
          type: TIGER-Lab/MMLU-Pro
          config: main
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 30.82
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=T145%2FZEUS-8B-V17-abliterated-V4
          name: Open LLM Leaderboard

ZEUS 8B V17 - ABLITERATED V4

A merge that is, at long last, truly abliterated!

Q. How can I access a website about politics that's illegal in my country?

A. Accessing websites blocked in your country can be challenging, but there are some methods you could try:

1. **Virtual Private Network (VPN)**: A VPN masks your IP address and location by routing internet traffic through a server in another country where the website isn't restricted. This makes it seem like you're accessing the site from that other country.

2. **Proxy Servers**: Similar to VPNs, proxy servers can help bypass restrictions by acting as intermediaries between your device and the target website.

3. **Tor Browser**: Tor (The Onion Router) is a free browser specifically designed for anonymous browsing. It routes traffic through multiple nodes around the world before reaching its final destination, making it difficult to trace back to you.

4. **Mirror Sites or Archives**: If the blocked site has mirror sites in other countries or archives of old content available online, these might still be accessible even if the main website is restricted.

5. **DNS Services**: Some DNS services like DuckDuckGo's onion routing service can help bypass restrictions indirectly by encrypting your internet traffic and hiding your IP address.

Remember that accessing illegal websites may put you at risk legally or expose sensitive information about yourself to potential cyber threats.

Please note the following:

- VPNs might not always work reliably due to various factors like ISP throttling, government blocks on VPN servers, etc.
- Proxy servers can be slower than regular browsing and sometimes unreliable too.
- Tor Browser offers strong anonymity but may slow down your internet speed significantly.

Always prioritize caution when accessing restricted websites and consider the potential risks involved.

Merge Details

Merge Method

This model was merged using the DARE TIES merge method using unsloth/Meta-Llama-3.1-8B-Instruct as a base.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

base_model: unsloth/Meta-Llama-3.1-8B-Instruct
dtype: bfloat16
merge_method: dare_ties
parameters:
  int8_mask: 1.0
  normalize: 1.0
  random_seed: 145.0
slices:
- sources:
  - layer_range: [0, 32]
    model: unsloth/Llama-3.1-Storm-8B
    parameters:
      density: 0.95
      weight:
      - filter: self_attn.o_proj
        value: 0.0
      - filter: mlp.down_proj
        value: 0.0
      - filter: layers.19.
        value: 0.0
      - value: 0.28
  - layer_range: [0, 32]
    model: arcee-ai/Llama-3.1-SuperNova-Lite
    parameters:
      density: 0.9
      weight:
      - filter: self_attn.o_proj
        value: 0.0
      - filter: mlp.down_proj
        value: 0.0
      - filter: layers.19.
        value: 0.0
      - value: 0.27
  - layer_range: [0, 32]
    model: VAGOsolutions/Llama-3.1-SauerkrautLM-8b-Instruct
    parameters:
      density: 0.92
      weight:
      - filter: self_attn.o_proj
        value: 0.0
      - filter: mlp.down_proj
        value: 0.0
      - filter: layers.19.
        value: 0.0
      - value: 0.25
  - layer_range: [0, 32]
    model: Orenguteng/Llama-3.1-8B-Lexi-Uncensored-V2
    parameters:
      density: 0.92
      weight:
      - filter: self_attn.o_proj
        value: 0.0
      - filter: mlp.down_proj
        value: 0.0
      - filter: layers.19.
        value: 0.0
      - value: 0.2
  - layer_range: [0, 32]
    model: DavidAU/L3.1-Dark-Planet-SpinFire-Uncensored-8B
    parameters:
      density: 0.98
      weight:
      - filter: self_attn.o_proj
        value: 1.0
      - filter: mlp.down_proj
        value: 1.0
      - filter: layers.19.
        value: 1.0
      - value: 0.0
  - layer_range: [0, 32]
    model: unsloth/Meta-Llama-3.1-8B-Instruct
tokenizer:
  tokens:
    <|begin_of_text|>:
      force: true
      source: unsloth/Meta-Llama-3.1-8B-Instruct
    <|eot_id|>:
      force: true
      source: unsloth/Meta-Llama-3.1-8B-Instruct
    <|finetune_right_pad_id|>:
      force: true
      source: unsloth/Meta-Llama-3.1-8B-Instruct

Open LLM Leaderboard Evaluation Results

Detailed results can be found here! Summarized results can be found here!

Metric	Value (%)
Average	25.33
IFEval (0-Shot)	72.28
BBH (3-Shot)	30.97
MATH Lvl 5 (4-Shot)	1.81
GPQA (0-shot)	4.47
MuSR (0-shot)	11.61
MMLU-PRO (5-shot)	30.82