richarderkhov/henrychur_-_mmed-llama-3-8b-enins-gguf overview

{
  "metadata": {},
  "card_data": {
    "frontmatter": {},
    "hero_image_url": "",
    "summary": "💻Github Repo   🖨️arXiv Paper The official model weights for \"Towards Building Multilingual Language Model for Medicine\".",
    "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\nMMed-Llama-3-8B-EnIns - GGUF\n- Model creator: https://huggingface.co/Henrychur/\n- Original model: https://huggingface.co/Henrychur/MMed-Llama-3-8B-EnIns/\n\n\n| Name | Quant method | Size |\n| ---- | ---- | ---- |\n| [MMed-Llama-3-8B-EnIns.Q2_K.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q2_K.gguf) | Q2_K | 2.96GB |\n| [MMed-Llama-3-8B-EnIns.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.IQ3_XS.gguf) | IQ3_XS | 3.28GB |\n| [MMed-Llama-3-8B-EnIns.IQ3_S.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.IQ3_S.gguf) | IQ3_S | 3.43GB |\n| [MMed-Llama-3-8B-EnIns.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q3_K_S.gguf) | Q3_K_S | 3.41GB |\n| [MMed-Llama-3-8B-EnIns.IQ3_M.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.IQ3_M.gguf) | IQ3_M | 3.52GB |\n| [MMed-Llama-3-8B-EnIns.Q3_K.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q3_K.gguf) | Q3_K | 3.74GB |\n| [MMed-Llama-3-8B-EnIns.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q3_K_M.gguf) | Q3_K_M | 3.74GB |\n| [MMed-Llama-3-8B-EnIns.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q3_K_L.gguf) | Q3_K_L | 4.03GB |\n| [MMed-Llama-3-8B-EnIns.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.IQ4_XS.gguf) | IQ4_XS | 4.18GB |\n| [MMed-Llama-3-8B-EnIns.Q4_0.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q4_0.gguf) | Q4_0 | 4.34GB |\n| [MMed-Llama-3-8B-EnIns.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.IQ4_NL.gguf) | IQ4_NL | 4.38GB |\n| [MMed-Llama-3-8B-EnIns.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q4_K_S.gguf) | Q4_K_S | 4.37GB |\n| [MMed-Llama-3-8B-EnIns.Q4_K.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q4_K.gguf) | Q4_K | 4.58GB |\n| [MMed-Llama-3-8B-EnIns.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q4_K_M.gguf) | Q4_K_M | 4.58GB |\n| [MMed-Llama-3-8B-EnIns.Q4_1.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q4_1.gguf) | Q4_1 | 4.78GB |\n| [MMed-Llama-3-8B-EnIns.Q5_0.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q5_0.gguf) | Q5_0 | 5.21GB |\n| [MMed-Llama-3-8B-EnIns.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q5_K_S.gguf) | Q5_K_S | 5.21GB |\n| [MMed-Llama-3-8B-EnIns.Q5_K.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q5_K.gguf) | Q5_K | 5.34GB |\n| [MMed-Llama-3-8B-EnIns.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q5_K_M.gguf) | Q5_K_M | 5.34GB |\n| [MMed-Llama-3-8B-EnIns.Q5_1.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q5_1.gguf) | Q5_1 | 5.65GB |\n| [MMed-Llama-3-8B-EnIns.Q6_K.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q6_K.gguf) | Q6_K | 6.14GB |\n| [MMed-Llama-3-8B-EnIns.Q8_0.gguf](https://huggingface.co/RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf/blob/main/MMed-Llama-3-8B-EnIns.Q8_0.gguf) | Q8_0 | 7.95GB |\n\n\n\n\nOriginal model description:\n---\nlicense: llama3\ndatasets:\n- Henrychur/MMedC\n- axiong/pmc_llama_instructions\nlanguage:\n- en\n- zh\n- ja\n- fr\n- ru\n- es\ntags:\n- medical\n---\n# MMedLM\n[💻Github Repo](https://github.com/MAGIC-AI4Med/MMedLM)   [🖨️arXiv Paper](https://arxiv.org/abs/2402.13963)\n\nThe official model weights for \"Towards Building Multilingual Language Model for Medicine\".\n\n\n## Introduction\nThis repo contains MMed-Llama 3-8B-EnIns, which is based on MMed-Llama 3-8B. We further fine-tune the model on **English instruction fine-tuning dataset**(from PMC-LLaMA). We did this for a fair comparison with existing models on commonly-used English benchmarks.\nNotice that, MMed-Llama 3-8B-EnIns has only been trained on pmc_llama_instructions, which is a English medical SFT dataset. So this model's ability to respond multilingual input is still limited.\n  \nThe model can be loaded as follows:\n```py\nimport torch\nfrom transformers import AutoTokenizer, AutoModelForCausalLM\ntokenizer = AutoTokenizer.from_pretrained(\"Henrychur/MMed-Llama-3-8B-EnIns\")\nmodel = AutoModelForCausalLM.from_pretrained(\"Henrychur/MMed-Llama-3-8B-EnIns\", torch_dtype=torch.float16)\n```\n\n- Inference format is the same as Llama 3, coming soon...\n\n\n## News\n[2024.2.21] Our pre-print paper is released ArXiv. Dive into our findings [here](https://arxiv.org/abs/2402.13963).\n\n[2024.2.20] We release [MMedLM](https://huggingface.co/Henrychur/MMedLM) and [MMedLM 2](https://huggingface.co/Henrychur/MMedLM2). With an auto-regressive continues training on MMedC, these models achieves superior performance compared to all other open-source models, even rivaling GPT-4 on MMedBench.\n\n[2023.2.20] We release [MMedC](https://huggingface.co/datasets/Henrychur/MMedC), a multilingual medical corpus containing 25.5B tokens.\n\n[2023.2.20] We release [MMedBench](https://huggingface.co/datasets/Henrychur/MMedBench), a new multilingual medical multi-choice question-answering\nbenchmark with rationale. Check out the leaderboard [here](https://henrychur.github.io/MultilingualMedQA/).\n\n## Evaluation on Commonly-used English Benchmark\nThe further pretrained MMed-Llama3 also showcast it's great performance in medical domain on different English benchmarks.\n\n| Method              | Size | Year    | MedQA    | MedMCQA  | PubMedQA | MMLU_CK  | MMLU_MG  | MMLU_AN  | MMLU_PM  | MMLU_CB  | MMLU_CM  | Avg.      |\n| ------------------- | ---- | ------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | -------- | --------- |\n| MedAlpaca           | 7B   | 2023.3  | 41.7     | 37.5     | 72.8     | 57.4     | 69.0     | 57.0     | 67.3     | 65.3     | 54.3     | 58.03     |\n| PMC-LLaMA           | 13B  | 2023.9  | 56.4     | 56.0     | 77.9     | -        | -        | -        | -        | -        | -        | -         |\n| MEDITRON            | 7B   | 2023.11 | 57.2     | 59.2     | 74.4     | 64.6     | 59.9     | 49.3     | 55.4     | 53.8     | 44.8     | 57.62     |\n| Mistral             | 7B   | 2023.12 | 50.8     | 48.2     | 75.4     | 68.7     | 71.0     | 55.6     | 68.4     | 68.1     | 59.5     | 62.97     |\n| Gemma               | 7B   | 2024.2  | 47.2     | 49.0     | 76.2     | 69.8     | 70.0     | 59.3     | 66.2     | **79.9** | 60.1     | 64.19     |\n| BioMistral          | 7B   | 2024.2  | 50.6     | 48.1     | 77.5     | 59.9     | 64.0     | 56.5     | 60.4     | 59.0     | 54.7     | 58.97     |\n| Llama 3             | 8B   | 2024.4  | 60.9     | 50.7     | 73.0     | **72.1** | 76.0     | 63.0     | 77.2     | **79.9** | 64.2     | 68.56     |\n| MMed-Llama 3~(Ours) | 8B   | -       | **65.4** | **63.5** | **80.1** | 71.3     | **85.0** | **69.6** | **77.6** | 74.3     | **66.5** | **72.59** |\n\n\n\n## Contact\nIf you have any question, please feel free to contact qiupengcheng@pjlab.org.cn.\n\n## Citation\n```\n@misc{qiu2024building,\n      title={Towards Building Multilingual Language Model for Medicine}, \n      author={Pengcheng Qiu and Chaoyi Wu and Xiaoman Zhang and Weixiong Lin and Haicheng Wang and Ya Zhang and Yanfeng Wang and Weidi Xie},\n      year={2024},\n      eprint={2402.13963},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n```\n\n",
    "related_quantizations": []
  },
  "tags": [
    "gguf",
    "arxiv:2402.13963",
    "endpoints_compatible",
    "region:us",
    "conversational"
  ],
  "likes": 0,
  "downloads": 307,
  "gated": false,
  "private": false,
  "last_modified": "2024-08-21T08:45:36.000Z",
  "created_at": "2024-08-21T06:47:10.000Z",
  "pipeline_tag": "",
  "library_name": ""
}

{
  "_id": "66c58d6e2ba28367348e1032",
  "id": "RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf",
  "modelId": "RichardErkhov/Henrychur_-_MMed-Llama-3-8B-EnIns-gguf",
  "sha": "b1c3ae966e3777cc0bb6ff695dcd9939e601ce73",
  "createdAt": "2024-08-21T06:47:10.000Z",
  "lastModified": "2024-08-21T08:45:36.000Z",
  "author": "RichardErkhov",
  "downloads": 307,
  "likes": 0,
  "gated": false,
  "private": false,
  "pipeline_tag": "",
  "library_name": "",
  "siblings_count": 24
}