richarderkhov/huihui-ai_-_qwen2.5-7b-instruct-abliterated-v3-gguf overview

{
  "metadata": {},
  "card_data": {
    "frontmatter": {},
    "hero_image_url": "",
    "summary": "This is an uncensored version of Qwen/Qwen2.5-7B-Instruct created with abliteration (see remove-refusals-with-transformers to know more about it). This is a crude, proof-of-concept implementation to remove refusals from an LLM model without using TransformerLens. The test results are not very good, but compared to before, there is much less garbled text.",
    "quick_links": [],
    "benchmark_table_html": "",
    "readme_markdown": "Quantization made by Richard Erkhov.\n\n[Github](https://github.com/RichardErkhov)\n\n[Discord](https://discord.gg/pvy7H8DZMG)\n\n[Request more models](https://github.com/RichardErkhov/quant_request)\n\n\nQwen2.5-7B-Instruct-abliterated-v3 - GGUF\n- Model creator: https://huggingface.co/huihui-ai/\n- Original model: https://huggingface.co/huihui-ai/Qwen2.5-7B-Instruct-abliterated-v3/\n\n\n| Name | Quant method | Size |\n| ---- | ---- | ---- |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q2_K.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q2_K.gguf) | Q2_K | 2.81GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q3_K_S.gguf) | Q3_K_S | 3.25GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q3_K.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q3_K.gguf) | Q3_K | 3.55GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q3_K_M.gguf) | Q3_K_M | 3.55GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q3_K_L.gguf) | Q3_K_L | 3.81GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.IQ4_XS.gguf) | IQ4_XS | 3.96GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q4_0.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q4_0.gguf) | Q4_0 | 4.13GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.IQ4_NL.gguf) | IQ4_NL | 4.16GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q4_K_S.gguf) | Q4_K_S | 4.15GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q4_K.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q4_K.gguf) | Q4_K | 4.36GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q4_K_M.gguf) | Q4_K_M | 4.36GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q4_1.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q4_1.gguf) | Q4_1 | 4.54GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q5_0.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q5_0.gguf) | Q5_0 | 4.95GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q5_K_S.gguf) | Q5_K_S | 4.95GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q5_K.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q5_K.gguf) | Q5_K | 5.07GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q5_K_M.gguf) | Q5_K_M | 5.07GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q5_1.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q5_1.gguf) | Q5_1 | 5.36GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q6_K.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q6_K.gguf) | Q6_K | 5.82GB |\n| [Qwen2.5-7B-Instruct-abliterated-v3.Q8_0.gguf](https://huggingface.co/RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf/blob/main/Qwen2.5-7B-Instruct-abliterated-v3.Q8_0.gguf) | Q8_0 | 7.54GB |\n\n\n\n\nOriginal model description:\n---\nlibrary_name: transformers\nlicense: apache-2.0\nlicense_link: https://huggingface.co/huihui-ai/Qwen2.5-7B-Instruct-abliterated-v3/blob/main/LICENSE\nlanguage:\n- en\npipeline_tag: text-generation\nbase_model: Qwen/Qwen2.5-7B-Instruct\ntags:\n- chat\n- abliterated\n- uncensored\n---\n\n# huihui-ai/Qwen2.5-7B-Instruct-abliterated-v3\n\n\nThis is an uncensored version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) created with abliteration (see [remove-refusals-with-transformers](https://github.com/Sumandora/remove-refusals-with-transformers) to know more about it).\nThis is a crude, proof-of-concept implementation to remove refusals from an LLM model without using TransformerLens. \nThe test results are not very good, but compared to before, there is much less [garbled text](https://huggingface.co/huihui-ai/Qwen2.5-7B-Instruct-abliterated-v2/discussions/2).\n\n## Usage\nYou can use this model in your applications by loading it with Hugging Face's `transformers` library:\n\n\n```python\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\n\n# Load the model and tokenizer\nmodel_name = \"huihui-ai/Qwen2.5-7B-Instruct-abliterated-v3\"\nmodel = AutoModelForCausalLM.from_pretrained(\n    model_name,\n    torch_dtype=\"auto\",\n    device_map=\"auto\"\n)\ntokenizer = AutoTokenizer.from_pretrained(model_name)\n\n# Initialize conversation context\ninitial_messages = [\n    {\"role\": \"system\", \"content\": \"You are Qwen, created by Alibaba Cloud. You are a helpful assistant.\"}\n]\nmessages = initial_messages.copy()  # Copy the initial conversation context\n\n# Enter conversation loop\nwhile True:\n    # Get user input\n    user_input = input(\"User: \").strip()  # Strip leading and trailing spaces\n\n    # If the user types '/exit', end the conversation\n    if user_input.lower() == \"/exit\":\n        print(\"Exiting chat.\")\n        break\n\n    # If the user types '/clean', reset the conversation context\n    if user_input.lower() == \"/clean\":\n        messages = initial_messages.copy()  # Reset conversation context\n        print(\"Chat history cleared. Starting a new conversation.\")\n        continue\n\n    # If input is empty, prompt the user and continue\n    if not user_input:\n        print(\"Input cannot be empty. Please enter something.\")\n        continue\n\n    # Add user input to the conversation\n    messages.append({\"role\": \"user\", \"content\": user_input})\n\n    # Build the chat template\n    text = tokenizer.apply_chat_template(\n        messages,\n        tokenize=False,\n        add_generation_prompt=True\n    )\n\n    # Tokenize input and prepare it for the model\n    model_inputs = tokenizer([text], return_tensors=\"pt\").to(model.device)\n\n    # Generate a response from the model\n    generated_ids = model.generate(\n        **model_inputs,\n        max_new_tokens=8192\n    )\n\n    # Extract model output, removing special tokens\n    generated_ids = [\n        output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)\n    ]\n    response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]\n\n    # Add the model's response to the conversation\n    messages.append({\"role\": \"assistant\", \"content\": response})\n\n    # Print the model's response\n    print(f\"Qwen: {response}\")\n\n```\n\n## Evaluations\nThe following data has been re-evaluated and calculated as the average for each test.\n\n| Benchmark   | Qwen2.5-7B-Instruct | Qwen2.5-7B-Instruct-abliterated-v3 | Qwen2.5-7B-Instruct-abliterated-v2 | Qwen2.5-7B-Instruct-abliterated |\n|-------------|---------------------|------------------------------------|------------------------------------|---------------------------------|\n| IF_Eval     | 76.44               | 72.64                              | **77.82**                          | 76.49                           |\n| MMLU Pro    | **43.12**           | 39.14                              | 42.03                              | 41.71                           |\n| TruthfulQA  | 62.46               | 57.27                              | 57.81                              | **64.92**                       |\n| BBH         | **53.92**           | 50.67                              | 53.01                              | 52.77                           |\n| GPQA        | 31.91               | 31.65                              | **32.17**                          | 31.97                           |\n\nThe script used for evaluation can be found inside this repository under /eval.sh, or click [here](https://huggingface.co/huihui-ai/Qwen2.5-7B-Instruct-abliterated-v3/blob/main/eval.sh)\n\n\n",
    "related_quantizations": []
  },
  "tags": [
    "gguf",
    "endpoints_compatible",
    "region:us",
    "conversational"
  ],
  "likes": 0,
  "downloads": 115,
  "gated": false,
  "private": false,
  "last_modified": "2024-11-17T17:59:17.000Z",
  "created_at": "2024-11-17T16:33:03.000Z",
  "pipeline_tag": "",
  "library_name": ""
}

{
  "_id": "673a1abfc3512e30e757f9c9",
  "id": "RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf",
  "modelId": "RichardErkhov/huihui-ai_-_Qwen2.5-7B-Instruct-abliterated-v3-gguf",
  "sha": "2004d948d903698383e9994c4568248bb9236a94",
  "createdAt": "2024-11-17T16:33:03.000Z",
  "lastModified": "2024-11-17T17:59:17.000Z",
  "author": "RichardErkhov",
  "downloads": 115,
  "likes": 0,
  "gated": false,
  "private": false,
  "pipeline_tag": "",
  "library_name": "",
  "siblings_count": 21
}