quantfactory/llama3-8b-darkidol-2.3-uncensored-32k-gguf overview

{
  "metadata": {},
  "card_data": {
    "license": "llama3",
    "language": [
      "en"
    ],
    "tags": [
      "roleplay",
      "llama3",
      "sillytavern",
      "idol"
    ],
    "frontmatter": {},
    "hero_image_url": "https://lh7-rt.googleusercontent.com/docsz/AD_4nXeiuCm7c8lEwEJuRey9kiVZsRn2W-b4pWlu3-X534V3YmVuVc2ZL-NXg2RkzSOOS2JXGHutDuyyNAUtdJI65jGTo8jT9Y99tMi4H4MqL44Uc5QKG77B0d6-JfIkZHFaUA71-RtjyYZWVIhqsNZcx8-OMaA?key=xt3VSDoCbmTY7o-cwwOFwQ",
    "summary": "This is quantized version of aifeifei798/llama3-8B-DarkIdol-2.3-Uncensored-32K created using llama.cpp # Original Model Card # The final version of Llama 3.0 will be followed by the next iteration starting from Llama 3.1. # Special Thanks: # These are my own quantizations (updated almost daily). The difference with normal quantizations is that I quantize the output and embed tensors to f16. and the other tensors to 15_k,q6_k or q8_0. This creates models that are little or not degraded at all and have a smaller size. They run at about 3-6 t/sec on CPU only using llama.cpp And obviously faster on computers with potent GPUs # Model Description: The module combination has been readjusted to better fulfill various roles and has been adapted for mobile phones. !image/png",
    "quick_links": [],
    "benchmark_table_html": "",
    "readme_markdown": "\n---\n\nlicense: llama3\nlanguage:\n- en\ntags:\n  - roleplay\n  - llama3\n  - sillytavern\n  - idol\n\n---\n\n[![QuantFactory Banner](https://lh7-rt.googleusercontent.com/docsz/AD_4nXeiuCm7c8lEwEJuRey9kiVZsRn2W-b4pWlu3-X534V3YmVuVc2ZL-NXg2RkzSOOS2JXGHutDuyyNAUtdJI65jGTo8jT9Y99tMi4H4MqL44Uc5QKG77B0d6-JfIkZHFaUA71-RtjyYZWVIhqsNZcx8-OMaA?key=xt3VSDoCbmTY7o-cwwOFwQ)](https://hf.co/QuantFactory)\n\n\n# QuantFactory/llama3-8B-DarkIdol-2.3-Uncensored-32K-GGUF\nThis is quantized version of [aifeifei798/llama3-8B-DarkIdol-2.3-Uncensored-32K](https://huggingface.co/aifeifei798/llama3-8B-DarkIdol-2.3-Uncensored-32K) created using llama.cpp\n\n# Original Model Card\n\n# The final version of Llama 3.0 will be followed by the next iteration starting from Llama 3.1.\n# Special Thanks:\n - Lewdiculous's superb gguf version, thank you for your conscientious and responsible dedication.\n - https://huggingface.co/LWDCLS/llama3-8B-DarkIdol-2.3-Uncensored-32K-GGUF-IQ-Imatrix-Request\n - mradermacher's superb gguf version, thank you for your conscientious and responsible dedication.\n - https://huggingface.co/mradermacher/llama3-8B-DarkIdol-2.3-Uncensored-32K-i1-GGUF\n - https://huggingface.co/mradermacher/llama3-8B-DarkIdol-2.3-Uncensored-32K-GGUF\n\n# These are my own quantizations (updated almost daily).\nThe difference with normal quantizations is that I quantize the output and embed tensors to f16.\nand the other tensors to 15_k,q6_k or q8_0.\nThis creates models that are little or not degraded at all and have a smaller size.\nThey run at about 3-6 t/sec on CPU only using llama.cpp\nAnd obviously faster on computers with potent GPUs\n- the fast cat at [ZeroWw/llama3-8B-DarkIdol-2.3-Uncensored-32K-GGUF](https://huggingface.co/ZeroWw/llama3-8B-DarkIdol-2.2-Uncensored-32K-GGUF)\n\n# Model Description:\nThe module combination has been readjusted to better fulfill various roles and has been adapted for mobile phones.\n- Saving money(LLama 3)\n- only test en.\n- Input Models input text only. Output Models generate text and code only.\n- Uncensored\n- Quick response\n- The underlying model used is winglian/Llama-3-8b-64k-PoSE (The theoretical support is 64k, but I have only tested up to 32k. :)\n- A scholarly response akin to a thesis.(I tend to write songs extensively, to the point where one song almost becomes as detailed as a thesis. :)\n- DarkIdol:Roles that you can imagine and those that you cannot imagine.\n- Roleplay\n- Specialized in various role-playing scenarios\n- more look at test role. (https://huggingface.co/aifeifei798/llama3-8B-DarkIdol-1.2/tree/main/test) \n- more look at LM Studio presets (https://huggingface.co/aifeifei798/llama3-8B-DarkIdol-1.2/tree/main/config-presets)\n![image/png](https://huggingface.co/aifeifei798/llama3-8B-DarkIdol-2.3-Uncensored-32K/resolve/main/llama3-8B-DarkIdol-2.3-Uncensored-32K.png)\n\n## virtual idol Twitter\n- https://x.com/aifeifei799\n\n# Questions\n- The model's response results are for reference only, please do not fully trust them.\n\n\n# Stop Strings\n```python\n    stop = [\n      \"## Instruction:\",\n      \"### Instruction:\",\n      \"<|end_of_text|>\",\n      \"  //:\",\n      \"</s>\",\n      \"<3```\",\n      \"### Note:\",\n      \"### Input:\",\n      \"### Response:\",\n      \"### Emoticons:\"\n    ],\n```\n# Model Use\n- Koboldcpp https://github.com/LostRuins/koboldcpp\n- Since KoboldCpp is taking a while to update with the latest llama.cpp commits, I'll recommend this [fork](https://github.com/Nexesenex/kobold.cpp) if anyone has issues.\n- LM Studio https://lmstudio.ai/\n- Please test again using the Default LM Studio Windows preset.\n- llama.cpp https://github.com/ggerganov/llama.cpp\n- Backyard AI https://backyard.ai/\n- Meet Layla,Layla is an AI chatbot that runs offline on your device.No internet connection required.No censorship.Complete privacy.Layla Lite https://www.layla-network.ai/\n- Layla Lite llama3-8B-DarkIdol-1.1-Q4_K_S-imat.gguf https://huggingface.co/LWDCLS/llama3-8B-DarkIdol-2.3-Uncensored-32K/blob/main/llama3-8B-DarkIdol-2.3-Uncensored-32K-Q4_K_S-imat.gguf?download=true\n- more gguf at https://huggingface.co/LWDCLS/llama3-8B-DarkIdol-2.3-Uncensored-32K-GGUF-IQ-Imatrix-Request\n# character\n- https://character-tavern.com/\n- https://characterhub.org/\n- https://pygmalion.chat/\n- https://aetherroom.club/\n- https://backyard.ai/\n- Layla AI chatbot\n### If you want to use vision functionality:\n * You must use the latest versions of [Koboldcpp](https://github.com/Nexesenex/kobold.cpp).\n \n### To use the multimodal capabilities of this model and use **vision** you need to load the specified **mmproj** file, this can be found inside this model repo. [Llava MMProj](https://huggingface.co/Nitral-AI/Llama-3-Update-3.0-mmproj-model-f16)\n \n * You can load the **mmproj** by using the corresponding section in the interface:\n ![image/png](https://cdn-uploads.huggingface.co/production/uploads/65d4cf2693a0a3744a27536c/UX6Ubss2EPNAT3SKGMLe0.png)\n### Thank you:\n To the authors for their hard work, which has given me more options to easily create what I want. Thank you for your efforts.\n- Hastagaras\n- Gryphe\n- cgato\n- ChaoticNeutrals\n- mergekit\n- merge\n- transformers\n- llama\n- Nitral-AI\n- MLP-KTLim\n- rinna\n- hfl\n- Rupesh2\n- stephenlzc\n- theprint\n- Sao10K\n- turboderp\n- TheBossLevel123\n- winglian\n- .........\n---\n# llama3-8B-DarkIdol-2.3-Uncensored-32K\n\nThis is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).\n\n## Merge Details\n### Merge Method\n\nThis model was merged using the [Model Stock](https://arxiv.org/abs/2403.19522) merge method using ./llama3-8B-DarkIdol-2.3b as a base.\n\n### Configuration\n\nThe following YAML configuration was used to produce this model:\n\n```yaml\nmodels:\n  - model: Sao10K/L3-8B-Niitama-v1\n  - model: Hastagaras/Jamet-8B-L3-MK.V-Blackroot\n  - model: Nitral-AI/Hathor_Tahsin-L3-8B-v0.85\n  - model: turboderp/llama3-turbcat-instruct-8b\n  - model: winglian/Llama-3-8b-64k-PoSE\nmerge_method: model_stock\nbase_model: winglian/Llama-3-8b-64k-PoSE\ndtype: bfloat16\n\nmodels:\n  - model: maldv/badger-writer-llama-3-8b\n  - model: underwoods/writer-8b\n  - model: Gryphe/Pantheon-RP-1.0-8b-Llama-3\n  - model: vicgalle/Roleplay-Llama-3-8B\n  - model: cgato/TheSalt-RP-L3-8b-DPO-v0.3.2-e0.15.2\n  - model: ./llama3-8B-DarkIdol-2.3a\nmerge_method: model_stock\nbase_model: ./llama3-8B-DarkIdol-2.3a\ndtype: bfloat16\n\nmodels:\n  - model: Rupesh2/Meta-Llama-3-8B-abliterated\n  - model: Orenguteng/Llama-3-8B-LexiFun-Uncensored-V1\n  - model: Orenguteng/Llama-3-8B-Lexi-Uncensored\n  - model: theprint/Llama-3-8B-Lexi-Smaug-Uncensored\n  - model: vicgalle/Unsafe-Llama-3-8B  \n  - model: vicgalle/Configurable-Hermes-2-Pro-Llama-3-8B\n  - model: ./llama3-8B-DarkIdol-2.3b\nmerge_method: model_stock\nbase_model: ./llama3-8B-DarkIdol-2.3b\ndtype: bfloat16\n```\n\n",
    "related_quantizations": []
  },
  "tags": [
    "gguf",
    "roleplay",
    "llama3",
    "sillytavern",
    "idol",
    "en",
    "arxiv:2403.19522",
    "license:llama3",
    "endpoints_compatible",
    "region:us"
  ],
  "likes": 7,
  "downloads": 134,
  "gated": false,
  "private": false,
  "last_modified": "2024-10-30T04:47:08.000Z",
  "created_at": "2024-10-30T04:03:09.000Z",
  "pipeline_tag": "",
  "library_name": ""
}

{
  "_id": "6721affd4d2b86b116161ce1",
  "id": "QuantFactory/llama3-8B-DarkIdol-2.3-Uncensored-32K-GGUF",
  "modelId": "QuantFactory/llama3-8B-DarkIdol-2.3-Uncensored-32K-GGUF",
  "sha": "3e924c85e491372b52721a420d1c8412aef92c8c",
  "createdAt": "2024-10-30T04:03:09.000Z",
  "lastModified": "2024-10-30T04:47:08.000Z",
  "author": "QuantFactory",
  "downloads": 134,
  "likes": 7,
  "gated": false,
  "private": false,
  "pipeline_tag": "",
  "library_name": "",
  "siblings_count": 16
}