lewdiculous/kunocchini-7b-128k-test-gguf-imatrix overview

{
  "metadata": {},
  "card_data": {
    "base_model": [
      "SanjiWatsuki/Kunoichi-DPO-v2-7B",
      "Epiculous/Fett-uccine-Long-Noodle-7B-120k-Context"
    ],
    "library_name": "transformers",
    "tags": [
      "mistral",
      "quantized",
      "text-generation-inference",
      "merge",
      "mergekit"
    ],
    "pipeline_tag": "text-generation",
    "inference": false,
    "frontmatter": {
      "base_model": [
        "SanjiWatsuki/Kunoichi-DPO-v2-7B",
        "Epiculous/Fett-uccine-Long-Noodle-7B-120k-Context"
      ],
      "library_name": "transformers",
      "tags": [
        "mistral",
        "quantized",
        "text-generation-inference",
        "merge",
        "mergekit"
      ],
      "pipeline_tag": "text-generation",
      "inference": "false"
    },
    "hero_image_url": "https://cdn-uploads.huggingface.co/production/uploads/642265bc01c62c1e4102dc36/9obNSalcJqCilQwr_4ssM.jpeg",
    "summary": "# UPDATED: Please download the v2 files that are now available. The new IQ4_NL and IQ4_XS quants were also added. # What does \"Imatrix\" mean? It stands for **Importance Matrix**, a technique used to improve the quality of quantized models. The **Imatrix** is calculated based on calibration data, and it helps determine the importance of different model activations during the quantization process. The idea is to preserve the most important information during quantization, which can help reduce the loss of model performance. One of the benefits of using an Imatrix is that it can lead to better model performance, especially when the calibration data is diverse. More information: [[1]](https://github.com/ggerganov/llama.cpp/discussions/5006) [[2]](https://github.com/ggerganov/llama.cpp/discussions/5263#discussioncomment-8395384)",
    "quick_links": [],
    "benchmark_table_html": "",
    "readme_markdown": "---\nbase_model:\n- SanjiWatsuki/Kunoichi-DPO-v2-7B\n- Epiculous/Fett-uccine-Long-Noodle-7B-120k-Context\nlibrary_name: transformers\ntags:\n- mistral\n- quantized\n- text-generation-inference\n- merge\n- mergekit\npipeline_tag: text-generation\ninference: false\n\n---\n\n> [!TIP]\n> **Support:** <br>\n> My upload speeds have been cooked and unstable lately. <br>\n> Realistically I'd need to move to get a better provider. <br>\n> If you **want** and you are able to... <br>\n> [**You can support my various endeavors here (Ko-fi).**](https://ko-fi.com/Lewdiculous) <br>\n> I apologize for disrupting your experience.\n\n\n# **GGUF-Imatrix quantizations for [Kunocchini-7b-128k-test](https://huggingface.co/Test157t/Kunocchini-7b-128k-test/).**\n\n# UPDATED: Please download the v2 files that are now available. The new IQ4_NL and IQ4_XS quants were also added.\n\n# What does \"Imatrix\" mean?\n\nIt stands for **Importance Matrix**, a technique used to improve the quality of quantized models.\n\nThe **Imatrix** is calculated based on calibration data, and it helps determine the importance of different model activations during the quantization process. The idea is to preserve the most important information during quantization, which can help reduce the loss of model performance.\n\nOne of the benefits of using an Imatrix is that it can lead to better model performance, especially when the calibration data is diverse.\n\nMore information: [[1]](https://github.com/ggerganov/llama.cpp/discussions/5006) [[2]](https://github.com/ggerganov/llama.cpp/discussions/5263#discussioncomment-8395384)\n\n## *This has been my personal favourite and daily-driver role-play model for a while, so I decided to make new quantizations for it using the full F16-Imatrix data.*\n\nSillyTavern preset files are located [here](https://huggingface.co/Test157t/Kunocchini-7b-128k-test/tree/main/ST%20presets).\n\n*If you want any specific quantization to be added, feel free to ask.*\n\nAll credits belong to the [creator](https://huggingface.co/Test157t/).\n\n`Base⇢ GGUF(F16)⇢ GGUF(Quants)`\n\nThe new **IQ3_S** merged today has shown to be better than the old Q3_K_S, so I added that instead of the later. Only supported in `koboldcpp-1.59.1` or higher.\n\nUsing [llama.cpp](https://github.com/ggerganov/llama.cpp/)-[b2254](https://github.com/ggerganov/llama.cpp/releases/tag/b2254).\n\nFor --imatrix data, `imatrix-Kunocchini-7b-128k-test-F16.dat` was used.\n\n# Original model information:\n\nThanks to @Epiculous for the dope model/ help with llm backends and support overall.\n\nId like to also thank @kalomaze for the dope sampler additions to ST. \n\n@SanjiWatsuki Thank you very much for the help, and the model!\n\nST users can find the TextGenPreset in the folder labeled so.\n\n![image/jpeg](https://cdn-uploads.huggingface.co/production/uploads/642265bc01c62c1e4102dc36/9obNSalcJqCilQwr_4ssM.jpeg)\n\nThe following models were included in the merge:\n* [SanjiWatsuki/Kunoichi-DPO-v2-7B](https://huggingface.co/SanjiWatsuki/Kunoichi-DPO-v2-7B)\n* [Epiculous/Fett-uccine-Long-Noodle-7B-120k-Context](https://huggingface.co/Epiculous/Fett-uccine-Long-Noodle-7B-120k-Context)\n\n### Configuration\n\nThe following YAML configuration was used to produce this model:\n\n```yaml\nslices:\n  - sources:\n      - model: SanjiWatsuki/Kunoichi-DPO-v2-7B\n        layer_range: [0, 32]\n      - model: Epiculous/Fett-uccine-Long-Noodle-7B-120k-Context\n        layer_range: [0, 32]\nmerge_method: slerp\nbase_model: SanjiWatsuki/Kunoichi-DPO-v2-7B\nparameters:\n  t:\n    - filter: self_attn\n      value: [0, 0.5, 0.3, 0.7, 1]\n    - filter: mlp\n      value: [1, 0.5, 0.7, 0.3, 0]\n    - value: 0.5\ndtype: bfloat16\n```",
    "related_quantizations": []
  },
  "tags": [
    "transformers",
    "gguf",
    "mistral",
    "quantized",
    "text-generation-inference",
    "merge",
    "mergekit",
    "text-generation",
    "base_model:Epiculous/Fett-uccine-Long-Noodle-7B-120k-Context",
    "base_model:merge:Epiculous/Fett-uccine-Long-Noodle-7B-120k-Context",
    "base_model:SanjiWatsuki/Kunoichi-DPO-v2-7B",
    "base_model:merge:SanjiWatsuki/Kunoichi-DPO-v2-7B",
    "region:us",
    "conversational"
  ],
  "likes": 29,
  "downloads": 261,
  "gated": false,
  "private": false,
  "last_modified": "2024-05-04T14:44:40.000Z",
  "created_at": "2024-02-25T04:16:49.000Z",
  "pipeline_tag": "text-generation",
  "library_name": "transformers"
}

{
  "_id": "65dabf310733541e0607b7c7",
  "id": "Lewdiculous/Kunocchini-7b-128k-test-GGUF-Imatrix",
  "modelId": "Lewdiculous/Kunocchini-7b-128k-test-GGUF-Imatrix",
  "sha": "f77ff6b312e8cf8f8291b692b5aae026f8b45780",
  "createdAt": "2024-02-25T04:16:49.000Z",
  "lastModified": "2024-05-04T14:44:40.000Z",
  "author": "Lewdiculous",
  "downloads": 261,
  "likes": 29,
  "gated": false,
  "private": false,
  "pipeline_tag": "text-generation",
  "library_name": "transformers",
  "siblings_count": 25
}