richarderkhov/ethzanalytics_-_ai-msgbot-gpt2-xl-dialogue-gguf overview

{
  "metadata": {},
  "card_data": {
    "frontmatter": {},
    "hero_image_url": "",
    "summary": "GPT2-XL (~1.5 B parameters) trained on the Wizard of Wikipedia dataset for 40k steps with **33**/36 layers frozen using aitextgen. The resulting model was then **further fine-tuned** on the Daily Dialogues for 40k steps, with **34**/36 layers frozen. Designed for use with ai-msgbot to create an open-ended chatbot (of course, if other use cases arise, have at it).",
    "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\nai-msgbot-gpt2-XL-dialogue - GGUF\n- Model creator: https://huggingface.co/ethzanalytics/\n- Original model: https://huggingface.co/ethzanalytics/ai-msgbot-gpt2-XL-dialogue/\n\n\n| Name | Quant method | Size |\n| ---- | ---- | ---- |\n| [ai-msgbot-gpt2-XL-dialogue.Q2_K.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q2_K.gguf) | Q2_K | 0.84GB |\n| [ai-msgbot-gpt2-XL-dialogue.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.IQ3_XS.gguf) | IQ3_XS | 0.84GB |\n| [ai-msgbot-gpt2-XL-dialogue.IQ3_S.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.IQ3_S.gguf) | IQ3_S | 0.84GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q3_K_S.gguf) | Q3_K_S | 0.84GB |\n| [ai-msgbot-gpt2-XL-dialogue.IQ3_M.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.IQ3_M.gguf) | IQ3_M | 0.91GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q3_K.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q3_K.gguf) | Q3_K | 0.97GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q3_K_M.gguf) | Q3_K_M | 0.97GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q3_K_L.gguf) | Q3_K_L | 1.03GB |\n| [ai-msgbot-gpt2-XL-dialogue.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.IQ4_XS.gguf) | IQ4_XS | 0.9GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q4_0.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q4_0.gguf) | Q4_0 | 0.91GB |\n| [ai-msgbot-gpt2-XL-dialogue.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.IQ4_NL.gguf) | IQ4_NL | 0.91GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q4_K_S.gguf) | Q4_K_S | 1.04GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q4_K.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q4_K.gguf) | Q4_K | 1.11GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q4_K_M.gguf) | Q4_K_M | 1.11GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q4_1.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q4_1.gguf) | Q4_1 | 1.0GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q5_0.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q5_0.gguf) | Q5_0 | 1.09GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q5_K_S.gguf) | Q5_K_S | 1.15GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q5_K.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q5_K.gguf) | Q5_K | 1.29GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q5_K_M.gguf) | Q5_K_M | 1.29GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q5_1.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q5_1.gguf) | Q5_1 | 1.18GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q6_K.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q6_K.gguf) | Q6_K | 1.52GB |\n| [ai-msgbot-gpt2-XL-dialogue.Q8_0.gguf](https://huggingface.co/RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf/blob/main/ai-msgbot-gpt2-XL-dialogue.Q8_0.gguf) | Q8_0 | 1.63GB |\n\n\n\n\nOriginal model description:\n---\nlanguage:\n- en\ntags:\n- text-generation\n- gpt2\n- gpt\nlicense: mit\ndatasets:\n- natural_questions\nwidget:\n- text: \"Do you like my new haircut?\\nperson beta:\\n\\n\"  \n  example_title: \"haircut\"\n- text: \"I love to learn new things.. are you willing to teach me something?\\nperson beta:\\n\\n\"  \n  example_title: \"teaching\"\n- text: \"What's your favorite animal? Mine is the dog? \\nperson beta:\\n\\n\"\n  example_title: \"favorite\"\n- text: \"how much does it cost?\\nperson beta:\\n\\n\"\n  example_title: \"money\"\ninference:\n  parameters:\n    min_length: 2\n    max_length: 64\n    length_penalty: 0.6\n    no_repeat_ngram_size: 3\n    do_sample: True\n    top_p: 0.85\n    top_k: 10\n    repetition_penalty: 2.1\n---\n\n# ai-msgbot: GPT2-XL-dialogue\n\n\nGPT2-XL (~1.5 B parameters) trained on [the Wizard of Wikipedia dataset](https://parl.ai/projects/wizard_of_wikipedia/) for 40k steps with **33**/36 layers frozen using `aitextgen`. The resulting model was then **further fine-tuned** on the [Daily Dialogues](http://yanran.li/dailydialog) for 40k steps, with **34**/36 layers frozen.\n\n\nDesigned for use with [ai-msgbot](https://github.com/pszemraj/ai-msgbot) to create an open-ended chatbot (of course, if other use cases arise, have at it).\n\n\n## conversation data\n\nThe dataset was tokenized and fed to the model as a conversation between two speakers, whose names are below. This is relevant for writing prompts and filtering/extracting text from responses.\n\n`script_speaker_name` = `person alpha`\n\n`script_responder_name` = `person beta`\n\n## examples\n\n- the default inference API examples should work _okay_\n- an ideal test would be explicitly adding `person beta` into the prompt text the model is forced to respond to instead of adding onto the entered prompt.\n\n\n## citations \n```\n@inproceedings{dinan2019wizard,\n  author={Emily Dinan and Stephen Roller and Kurt Shuster and Angela Fan and Michael Auli and Jason Weston},\n  title={{W}izard of {W}ikipedia: Knowledge-powered Conversational Agents},\n  booktitle = {Proceedings of the International Conference on Learning Representations (ICLR)},\n  year={2019},\n}\n\n@inproceedings{li-etal-2017-dailydialog,\n    title = \"{D}aily{D}ialog: A Manually Labelled Multi-turn Dialogue Dataset\",\n    author = \"Li, Yanran  and\n      Su, Hui  and\n      Shen, Xiaoyu  and\n      Li, Wenjie  and\n      Cao, Ziqiang  and\n      Niu, Shuzi\",\n    booktitle = \"Proceedings of the Eighth International Joint Conference on Natural Language Processing (Volume 1: Long Papers)\",\n    month = nov,\n    year = \"2017\",\n    address = \"Taipei, Taiwan\",\n    publisher = \"Asian Federation of Natural Language Processing\",\n    url = \"https://aclanthology.org/I17-1099\",\n    pages = \"986--995\",\n    abstract = \"We develop a high-quality multi-turn dialog dataset, \\textbf{DailyDialog}, which is intriguing in several aspects. The language is human-written and less noisy. The dialogues in the dataset reflect our daily communication way and cover various topics about our daily life. We also manually label the developed dataset with communication intention and emotion information. Then, we evaluate existing approaches on DailyDialog dataset and hope it benefit the research field of dialog systems. The dataset is available on \\url{http://yanran.li/dailydialog}\",\n}\n```\n\n",
    "related_quantizations": []
  },
  "tags": [
    "gguf",
    "endpoints_compatible",
    "region:us"
  ],
  "likes": 0,
  "downloads": 753,
  "gated": false,
  "private": false,
  "last_modified": "2024-10-04T18:57:00.000Z",
  "created_at": "2024-10-04T17:22:26.000Z",
  "pipeline_tag": "",
  "library_name": ""
}

{
  "_id": "670024520ccb3ca95879670b",
  "id": "RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf",
  "modelId": "RichardErkhov/ethzanalytics_-_ai-msgbot-gpt2-XL-dialogue-gguf",
  "sha": "832a925c9fb09575c7dbee47ede7c76c2896a23f",
  "createdAt": "2024-10-04T17:22:26.000Z",
  "lastModified": "2024-10-04T18:57:00.000Z",
  "author": "RichardErkhov",
  "downloads": 753,
  "likes": 0,
  "gated": false,
  "private": false,
  "pipeline_tag": "",
  "library_name": "",
  "siblings_count": 24
}