richarderkhov/jsoo_-_llama3-beomi-open-ko-8b-instruct-preview-test6-gguf overview

{
  "metadata": {},
  "card_data": {
    "frontmatter": {},
    "hero_image_url": "",
    "summary": "",
    "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\nLlama3-beomi-Open-Ko-8B-Instruct-preview-test6 - GGUF\n- Model creator: https://huggingface.co/Jsoo/\n- Original model: https://huggingface.co/Jsoo/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6/\n\n\n| Name | Quant method | Size |\n| ---- | ---- | ---- |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q2_K.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q2_K.gguf) | Q2_K | 2.96GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.IQ3_XS.gguf) | IQ3_XS | 3.28GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.IQ3_S.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.IQ3_S.gguf) | IQ3_S | 3.43GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q3_K_S.gguf) | Q3_K_S | 3.41GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.IQ3_M.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.IQ3_M.gguf) | IQ3_M | 3.52GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q3_K.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q3_K.gguf) | Q3_K | 3.74GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q3_K_M.gguf) | Q3_K_M | 3.74GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q3_K_L.gguf) | Q3_K_L | 4.03GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.IQ4_XS.gguf) | IQ4_XS | 4.18GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q4_0.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q4_0.gguf) | Q4_0 | 4.34GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.IQ4_NL.gguf) | IQ4_NL | 4.38GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q4_K_S.gguf) | Q4_K_S | 4.37GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q4_K.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q4_K.gguf) | Q4_K | 4.58GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q4_K_M.gguf) | Q4_K_M | 4.58GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q4_1.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q4_1.gguf) | Q4_1 | 4.78GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q5_0.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q5_0.gguf) | Q5_0 | 5.21GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q5_K_S.gguf) | Q5_K_S | 5.21GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q5_K.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q5_K.gguf) | Q5_K | 5.34GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q5_K_M.gguf) | Q5_K_M | 5.34GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q5_1.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q5_1.gguf) | Q5_1 | 5.65GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q6_K.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q6_K.gguf) | Q6_K | 6.14GB |\n| [Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q8_0.gguf](https://huggingface.co/RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf/blob/main/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6.Q8_0.gguf) | Q8_0 | 7.95GB |\n\n\n\n\nOriginal model description:\n---\nlanguage:\n- en\n- ko\nlicense: other\ntags:\n- facebook\n- meta\n- pytorch\n- llama\n- llama-3\n- llama-3-ko\npipeline_tag: text-generation\nlicense_name: llama3\nlicense_link: LICENSE\n---\n\n# Model Card for Model ID\n\n\n\n\n## Model Details\n\nLlama-3-Open-Ko-8B model is continued pretrained language model based on Llama-3-8B.\n\nThis model is trained fully with publicily available resource, with 60GB+ of deduplicated texts.\n\nWith the new Llama-3 tokenizer, the pretraining conducted with 17.7B+ tokens, which slightly more than Korean tokenizer(Llama-2-Ko tokenizer).\n\n\n**Sample usage**\n\n```\n  from transformers import pipeline\n  import torch\n  \n  pipe = pipeline(\n      task=\"text-generation\",\n      model=model,\n      tokenizer=tokenizer,\n      model_kwargs={\"torch_dtype\": torch.bfloat16},\n      truncation=True\n  )\n  \n  def extract_response_llama3(question):\n      messages = [\n          {\"role\": \"system\", \"content\": \"\"},\n          {\"role\": \"user\", \"content\": question},\n      ]\n  \n      prompt = pipe.tokenizer.apply_chat_template(\n          messages,\n          tokenize=False,\n          add_generation_prompt=True\n      )\n  \n      terminators = [\n          pipe.tokenizer.eos_token_id,\n          pipe.tokenizer.convert_tokens_to_ids(\"<|eot_id|>\")\n      ]\n  \n      outputs = pipe(\n          prompt,\n          max_new_tokens=256,\n          eos_token_id=terminators,\n          do_sample=True,\n          temperature=0.1,\n          top_p=0.9,\n          num_return_sequences=1\n      )\n  \n      return outputs[0]['generated_text'].split('\\n')[-1]\n  \n  \n  question = \"예산을 분배할 때 사업의 우선 순위를 정해서 차등 지원하는 방법을 뭐라고 하지\"\n  response = extract_response_llama3(question)\n  print(response)\n  \n  question = \"미세먼지 생성물질의 배출을 저감하고 종합적으로 관리하기 위한 법을 어디서 제정했니\"\n  response = extract_response_llama3(question)\n  print(response)\n  \n  question = \"어떤 장소의 대기오염을 방지하기 위한 정책의 법적 근거가 특별법의 제정으로 준비되었지\"\n  response = extract_response_llama3(question)\n  print(response)\n```\n\n**Sample Output**\n\n```\n선택과 집중\n\n환경부\n\n항만\n```\n\n\n",
    "related_quantizations": []
  },
  "tags": [
    "gguf",
    "endpoints_compatible",
    "region:us",
    "conversational"
  ],
  "likes": 0,
  "downloads": 131,
  "gated": false,
  "private": false,
  "last_modified": "2024-08-21T18:32:15.000Z",
  "created_at": "2024-08-21T16:40:15.000Z",
  "pipeline_tag": "",
  "library_name": ""
}

{
  "_id": "66c6186f4224c7abdc7b5ba7",
  "id": "RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf",
  "modelId": "RichardErkhov/Jsoo_-_Llama3-beomi-Open-Ko-8B-Instruct-preview-test6-gguf",
  "sha": "87405662b83b32c483ab2667f6405321522bb65f",
  "createdAt": "2024-08-21T16:40:15.000Z",
  "lastModified": "2024-08-21T18:32:15.000Z",
  "author": "RichardErkhov",
  "downloads": 131,
  "likes": 0,
  "gated": false,
  "private": false,
  "pipeline_tag": "",
  "library_name": "",
  "siblings_count": 24
}