richarderkhov/jslin09_-_gemma2-2b-fraud-gguf overview

{
  "metadata": {},
  "card_data": {
    "frontmatter": {},
    "hero_image_url": "",
    "summary": "本模型是以司法院公開之「詐欺」案件判決書做成之資料集，基於 Google Gemma2:2b 模型進行微調訓練，可以自動生成詐欺及竊盜案件之犯罪事實段落之草稿。資料集之資料範圍從100年1月1日至110年12月31日，所蒐集到的原始資料共有 74823 篇（判決以及裁定），我們只取判決書的「犯罪事實」欄位內容，並把這原始的資料分成三份，用於訓練的資料集有59858篇，約佔原始資料的80%，剩下的20%，則是各分配10%給驗證集（7482篇），10%給測試集（7483篇）。在本網頁進行測試時，請在模型載入完畢並生成第一小句後，持續按下Compute按鈕，就能持續生成文字。或是輸入自己想要測試的資料到文字框中進行測試。或是可以到這裡有更完整的使用體驗。 # 使用範例 如果要在自己的程式中調用本模型，可以參考下列的 Python 程式碼，藉由呼叫 API 的方式來生成刑事判決書「犯罪事實」欄的內容。   點擊後展開    import requests, json from time import sleep from tqdm.auto import tqdm, trange API_URL = \"https://api-inference.huggingface.co/models/jslin09/gemma2-2b-fraud\" API_TOKEN = 'XXXXXXXXXXXXXXX' # 調用模型的 API token headers = {\"Authorization\": f\"Bearer {API_TOKEN}\"} def query(payload): response = requests.post(API_URL, headers=headers, json=payload) return json.loads(response.content.decode(\"utf-8\")) prompt = \"森上梅前明知其無資力支付酒店消費，亦無付款意願，竟意圖為自己不法之所有，\" query_dict = { \"inputs\": prompt, } text_len = 300 t = trange(text_len, desc= '生成例稿', leave=True) for i in t: response = query(query_dict) try: response_text = response[0]['generated_text'] query_dict[\"inputs\"] = response_text t.set_description(f\"{i}: {response[0]['generated_text']}\") t.refresh() except KeyError: sleep(30) # 如果伺服器太忙無回應，等30秒後再試。 pass print(response[0]['generated_text'])    或是，你要使用 transformers 套件來實作你的程式，將本模型下載至你本地端的電腦中執行，可以參考下列程式碼:   點擊後展開    # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained(\"jslin09/gemma2-2b-fraud\") model = AutoModelForCausalLM.from_pretrained(\"jslin09/gemma2-2b-fraud\")    # 致謝 微調本模型所需要的算力，是由評律網提供 NVIDIA H100。特此致謝。 # 引文訊息 `` @misc{lin2024legal, title={Legal Documents Drafting with Fine-Tuned Pre-Trained Large Language Model}, author={Chun-Hsien Lin and Pu-Jen Cheng}, year={2024}, eprint={2406.04202}, archivePrefix={arXiv}, primaryClass={cs.CL} url = {https://arxiv.org/abs/2406.04202} } ``",
    "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\ngemma2-2b-fraud - GGUF\n- Model creator: https://huggingface.co/jslin09/\n- Original model: https://huggingface.co/jslin09/gemma2-2b-fraud/\n\n\n| Name | Quant method | Size |\n| ---- | ---- | ---- |\n| [gemma2-2b-fraud.Q2_K.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q2_K.gguf) | Q2_K | 1.15GB |\n| [gemma2-2b-fraud.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q3_K_S.gguf) | Q3_K_S | 1.27GB |\n| [gemma2-2b-fraud.Q3_K.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q3_K.gguf) | Q3_K | 1.36GB |\n| [gemma2-2b-fraud.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q3_K_M.gguf) | Q3_K_M | 1.36GB |\n| [gemma2-2b-fraud.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q3_K_L.gguf) | Q3_K_L | 1.44GB |\n| [gemma2-2b-fraud.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.IQ4_XS.gguf) | IQ4_XS | 1.47GB |\n| [gemma2-2b-fraud.Q4_0.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q4_0.gguf) | Q4_0 | 1.52GB |\n| [gemma2-2b-fraud.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.IQ4_NL.gguf) | IQ4_NL | 1.53GB |\n| [gemma2-2b-fraud.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q4_K_S.gguf) | Q4_K_S | 1.53GB |\n| [gemma2-2b-fraud.Q4_K.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q4_K.gguf) | Q4_K | 1.59GB |\n| [gemma2-2b-fraud.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q4_K_M.gguf) | Q4_K_M | 1.59GB |\n| [gemma2-2b-fraud.Q4_1.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q4_1.gguf) | Q4_1 | 1.64GB |\n| [gemma2-2b-fraud.Q5_0.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q5_0.gguf) | Q5_0 | 1.75GB |\n| [gemma2-2b-fraud.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q5_K_S.gguf) | Q5_K_S | 1.75GB |\n| [gemma2-2b-fraud.Q5_K.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q5_K.gguf) | Q5_K | 1.79GB |\n| [gemma2-2b-fraud.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q5_K_M.gguf) | Q5_K_M | 1.79GB |\n| [gemma2-2b-fraud.Q5_1.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q5_1.gguf) | Q5_1 | 1.87GB |\n| [gemma2-2b-fraud.Q6_K.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q6_K.gguf) | Q6_K | 2.0GB |\n| [gemma2-2b-fraud.Q8_0.gguf](https://huggingface.co/RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf/blob/main/gemma2-2b-fraud.Q8_0.gguf) | Q8_0 | 2.59GB |\n\n\n\n\nOriginal model description:\n---\nlicense: gemma\ndatasets:\n- jslin09/Fraud_Case_Verdicts\nlanguage:\n- zh\nbase_model:\n- google/gemma-2-2b\npipeline_tag: text-generation\ntext-generation:\n  parameters:\n    max_length: 400\n    max_new_tokens: 400\n    do_sample: true\n    temperature: 0.75\n    top_k: 50\n    top_p: 0.9\ntags:\n- legal\nwidget:\n- text: 王大明意圖為自己不法所有，基於竊盜之犯意，\n  example_title: 生成竊盜罪之犯罪事實\n- text: 騙人布意圖為自己不法所有，基於詐欺取財之犯意，\n  example_title: 生成詐欺罪之犯罪事實\n- text: 梅友乾明知其無資力支付酒店消費，亦無付款意願，竟意圖為自己不法之所有，\n  example_title: 生成吃霸王餐之詐欺犯罪事實\n- text: 闕很大明知金融帳戶之存摺、提款卡及密碼係供自己使用之重要理財工具，\n  example_title: 生成賣帳戶幫助詐欺犯罪事實\n- text: 通訊王明知近來盛行以虛設、租賃、借用或買賣行動電話人頭門號之方式，供詐騙集團作為詐欺他人交付財物等不法用途，\n  example_title: 生成賣電話SIM卡之幫助詐欺犯罪事實\n- text: 趙甲王基於行使偽造特種文書及詐欺取財之犯意，\n  example_title: 偽造特種文書(契約、車牌等)詐財\nlibrary_name: transformers\n---\n# 判決書「犯罪事實」欄草稿自動生成\n本模型是以司法院公開之「詐欺」案件判決書做成之資料集，基於 [Google Gemma2:2b](https://huggingface.co/google/gemma-2-2b) 模型進行微調訓練，可以自動生成詐欺及竊盜案件之犯罪事實段落之草稿。資料集之資料範圍從100年1月1日至110年12月31日，所蒐集到的原始資料共有 74823 篇（判決以及裁定），我們只取判決書的「犯罪事實」欄位內容，並把這原始的資料分成三份，用於訓練的資料集有59858篇，約佔原始資料的80%，剩下的20%，則是各分配10%給驗證集（7482篇），10%給測試集（7483篇）。在本網頁進行測試時，請在模型載入完畢並生成第一小句後，持續按下Compute按鈕，就能持續生成文字。或是輸入自己想要測試的資料到文字框中進行測試。或是可以到[這裡](https://huggingface.co/spaces/jslin09/legal_document_drafting)有更完整的使用體驗。\n\n# 使用範例\n如果要在自己的程式中調用本模型，可以參考下列的 Python 程式碼，藉由呼叫 API 的方式來生成刑事判決書「犯罪事實」欄的內容。\n<details>\n  <summary> 點擊後展開 </summary>\n<pre>\n  <code>\nimport requests, json\nfrom time import sleep\nfrom tqdm.auto import tqdm, trange\n\nAPI_URL = \"https://api-inference.huggingface.co/models/jslin09/gemma2-2b-fraud\"\nAPI_TOKEN = 'XXXXXXXXXXXXXXX' # 調用模型的 API token\nheaders = {\"Authorization\": f\"Bearer {API_TOKEN}\"} \n\ndef query(payload):\n    response = requests.post(API_URL, headers=headers, json=payload)\n    return json.loads(response.content.decode(\"utf-8\"))\n\nprompt = \"森上梅前明知其無資力支付酒店消費，亦無付款意願，竟意圖為自己不法之所有，\"\nquery_dict = {\n\t\"inputs\": prompt,\n}\ntext_len = 300\nt = trange(text_len, desc= '生成例稿', leave=True)\nfor i in t:\n    response = query(query_dict)\n    try:\n        response_text = response[0]['generated_text']\n        query_dict[\"inputs\"] = response_text\n        t.set_description(f\"{i}: {response[0]['generated_text']}\")\n        t.refresh()\n    except KeyError:\n        sleep(30) # 如果伺服器太忙無回應，等30秒後再試。\n        pass\nprint(response[0]['generated_text'])\n</code>\n</pre>\n</details>\n\n或是，你要使用 transformers 套件來實作你的程式，將本模型下載至你本地端的電腦中執行，可以參考下列程式碼:\n<details>\n  <summary> 點擊後展開 </summary>\n<pre>\n  <code>\n# Load model directly\nfrom transformers import AutoTokenizer, AutoModelForCausalLM\n\ntokenizer = AutoTokenizer.from_pretrained(\"jslin09/gemma2-2b-fraud\")\nmodel = AutoModelForCausalLM.from_pretrained(\"jslin09/gemma2-2b-fraud\")\n\n</code>\n</pre>\n</details>\n\n# 致謝\n微調本模型所需要的算力，是由[評律網](https://www.pingluweb.com.tw/)提供 NVIDIA H100。特此致謝。\n\n# 引文訊息\n\n```\n@misc{lin2024legal,\n      title={Legal Documents Drafting with Fine-Tuned Pre-Trained Large Language Model}, \n      author={Chun-Hsien Lin and Pu-Jen Cheng},\n      year={2024},\n      eprint={2406.04202},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n      url = {https://arxiv.org/abs/2406.04202}\n}\n```\n\n",
    "related_quantizations": []
  },
  "tags": [
    "gguf",
    "arxiv:2406.04202",
    "endpoints_compatible",
    "region:us"
  ],
  "likes": 0,
  "downloads": 142,
  "gated": false,
  "private": false,
  "last_modified": "2024-10-29T22:26:35.000Z",
  "created_at": "2024-10-29T20:50:10.000Z",
  "pipeline_tag": "",
  "library_name": ""
}

{
  "_id": "67214a822c9459ac870dfba2",
  "id": "RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf",
  "modelId": "RichardErkhov/jslin09_-_gemma2-2b-fraud-gguf",
  "sha": "7dd91f232468d9641ac85a9b5d7c55f601e7a348",
  "createdAt": "2024-10-29T20:50:10.000Z",
  "lastModified": "2024-10-29T22:26:35.000Z",
  "author": "RichardErkhov",
  "downloads": 142,
  "likes": 0,
  "gated": false,
  "private": false,
  "pipeline_tag": "",
  "library_name": "",
  "siblings_count": 21
}