richarderkhov/kyujinpy_-_korean-openorca-v3-gguf overview

{
  "metadata": {},
  "card_data": {
    "frontmatter": {},
    "hero_image_url": "./Korean-OpenOrca.png",
    "summary": "!img",
    "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\nKorean-OpenOrca-v3 - GGUF\n- Model creator: https://huggingface.co/kyujinpy/\n- Original model: https://huggingface.co/kyujinpy/Korean-OpenOrca-v3/\n\n\n| Name | Quant method | Size |\n| ---- | ---- | ---- |\n| [Korean-OpenOrca-v3.Q2_K.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q2_K.gguf) | Q2_K | 4.52GB |\n| [Korean-OpenOrca-v3.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.IQ3_XS.gguf) | IQ3_XS | 4.99GB |\n| [Korean-OpenOrca-v3.IQ3_S.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.IQ3_S.gguf) | IQ3_S | 5.27GB |\n| [Korean-OpenOrca-v3.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q3_K_S.gguf) | Q3_K_S | 5.27GB |\n| [Korean-OpenOrca-v3.IQ3_M.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.IQ3_M.gguf) | IQ3_M | 5.57GB |\n| [Korean-OpenOrca-v3.Q3_K.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q3_K.gguf) | Q3_K | 5.9GB |\n| [Korean-OpenOrca-v3.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q3_K_M.gguf) | Q3_K_M | 5.9GB |\n| [Korean-OpenOrca-v3.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q3_K_L.gguf) | Q3_K_L | 6.45GB |\n| [Korean-OpenOrca-v3.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.IQ4_XS.gguf) | IQ4_XS | 6.54GB |\n| [Korean-OpenOrca-v3.Q4_0.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q4_0.gguf) | Q4_0 | 6.86GB |\n| [Korean-OpenOrca-v3.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.IQ4_NL.gguf) | IQ4_NL | 6.9GB |\n| [Korean-OpenOrca-v3.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q4_K_S.gguf) | Q4_K_S | 6.91GB |\n| [Korean-OpenOrca-v3.Q4_K.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q4_K.gguf) | Q4_K | 7.33GB |\n| [Korean-OpenOrca-v3.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q4_K_M.gguf) | Q4_K_M | 7.33GB |\n| [Korean-OpenOrca-v3.Q4_1.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q4_1.gguf) | Q4_1 | 7.61GB |\n| [Korean-OpenOrca-v3.Q5_0.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q5_0.gguf) | Q5_0 | 8.36GB |\n| [Korean-OpenOrca-v3.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q5_K_S.gguf) | Q5_K_S | 8.36GB |\n| [Korean-OpenOrca-v3.Q5_K.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q5_K.gguf) | Q5_K | 8.6GB |\n| [Korean-OpenOrca-v3.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q5_K_M.gguf) | Q5_K_M | 8.6GB |\n| [Korean-OpenOrca-v3.Q5_1.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q5_1.gguf) | Q5_1 | 9.1GB |\n| [Korean-OpenOrca-v3.Q6_K.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q6_K.gguf) | Q6_K | 9.95GB |\n| [Korean-OpenOrca-v3.Q8_0.gguf](https://huggingface.co/RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf/blob/main/Korean-OpenOrca-v3.Q8_0.gguf) | Q8_0 | 12.88GB |\n\n\n\n\nOriginal model description:\n---\nlanguage:\n- ko\ndatasets:\n- kyujinpy/OpenOrca-ko-v3\nlibrary_name: transformers\npipeline_tag: text-generation\nlicense: cc-by-nc-sa-4.0\n---\n**(주)미디어그룹사람과숲과 (주)마커의 LLM 연구 컨소시엄에서 개발된 모델입니다**  \n**The license is `cc-by-nc-sa-4.0`.**  \n\n# **🐳Korean-OpenOrca-13B-v2🐳**  \n![img](./Korean-OpenOrca.png)  \n\n## Model Details\n\n**Model Developers** Kyujin Han (kyujinpy)\n\n**Model Architecture**  \nKorean-OpenOrca-13B is an auto-regressive language model based on the LLaMA2 transformer architecture.\n\n**Repo Link**  \nGithub Korean-OpenOrca: [🐳Korean-OpenOrca🐳](https://github.com/Marker-Inc-Korea/Korean-OpenOrca)  \n\n**Base Model**  [hyunseoki/ko-en-llama2-13b](https://huggingface.co/hyunseoki/ko-en-llama2-13b)   \n\n**Training Dataset**  \nI use [OpenOrca-ko-v3](https://huggingface.co/datasets/kyujinpy/OpenOrca-ko-v3).  \nUsing DeepL, translate about [OpenOrca](https://huggingface.co/datasets/Open-Orca/OpenOrca).\n\nI use A100 GPU 40GB and COLAB, when trianing.\n\n# Model comparisons\n| Model | Average |Ko-ARC | Ko-HellaSwag | Ko-MMLU | Ko-TruthfulQA | Ko-CommonGen V2 |\n| --- | --- | --- | --- | --- | --- | --- |\n| [Korean-OpenOrca-13B🐳] | 48.79 | 43.09 | 54.13 | 40.24 | 45.22 | 61.28 |  \n| [Korean-OpenOrca-13B-v2🐳] | 48.17 | 43.17 | 54.51 | 42.90 | 41.82 | 58.44 | \n| Korean-OpenOrca-13B-v3🐳 | 48.86 | 43.77 | 54.30 | 41.79 | 43.85 | 60.57 | \n\n# Implementation Code\n```python\n### KO-Platypus\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\nimport torch\n\nrepo = \"kyujinpy/Korean-OpenOrca-13B-v3\"\nOpenOrca = AutoModelForCausalLM.from_pretrained(\n        repo,\n        return_dict=True,\n        torch_dtype=torch.float16,\n        device_map='auto'\n)\nOpenOrca_tokenizer = AutoTokenizer.from_pretrained(repo)\n```\n\n---\n\n",
    "related_quantizations": []
  },
  "tags": [
    "gguf",
    "endpoints_compatible",
    "region:us"
  ],
  "likes": 0,
  "downloads": 169,
  "gated": false,
  "private": false,
  "last_modified": "2024-10-10T01:06:02.000Z",
  "created_at": "2024-10-09T20:17:51.000Z",
  "pipeline_tag": "",
  "library_name": ""
}

{
  "_id": "6706e4efd6f7382cc40b8015",
  "id": "RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf",
  "modelId": "RichardErkhov/kyujinpy_-_Korean-OpenOrca-v3-gguf",
  "sha": "4d3bbf3b8c4c14fb314cd49a154f36de860ac3c1",
  "createdAt": "2024-10-09T20:17:51.000Z",
  "lastModified": "2024-10-10T01:06:02.000Z",
  "author": "RichardErkhov",
  "downloads": 169,
  "likes": 0,
  "gated": false,
  "private": false,
  "pipeline_tag": "",
  "library_name": "",
  "siblings_count": 24
}