richarderkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf overview

{
  "metadata": {},
  "card_data": {
    "frontmatter": {},
    "hero_image_url": "",
    "summary": "SEA-LION is a collection of Large Language Models (LLMs) which has been pretrained and instruct-tuned for the Southeast Asia (SEA) region. Llama3 8B CPT SEA-Lionv2.1 Instruct is a multilingual model which has been fine-tuned with around **100,000 English instruction-completion pairs** alongside a smaller pool of around **50,000 instruction-completion pairs** from other ASEAN languages, such as Indonesian, Thai and Vietnamese. These instructions have been carefully curated and rewritten to ensure the model was trained on truly open, commercially permissive and high quality datasets. Llama3 8B CPT SEA-Lionv2.1 Instruct has undergone additional supervised fine-tuning and alignment compared to the now deprecated Llama3 8B CPT SEA-Lionv2 Instruct. These improvements have increased the model's capabilities in chat interactions and its ability to follow instructions accurately. SEA-LION stands for _Southeast Asian Languages In One Network_.",
    "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\nllama3-8b-cpt-sea-lionv2.1-instruct - GGUF\n- Model creator: https://huggingface.co/aisingapore/\n- Original model: https://huggingface.co/aisingapore/llama3-8b-cpt-sea-lionv2.1-instruct/\n\n\n| Name | Quant method | Size |\n| ---- | ---- | ---- |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q2_K.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q2_K.gguf) | Q2_K | 2.96GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.IQ3_XS.gguf) | IQ3_XS | 3.28GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.IQ3_S.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.IQ3_S.gguf) | IQ3_S | 3.43GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q3_K_S.gguf) | Q3_K_S | 3.41GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.IQ3_M.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.IQ3_M.gguf) | IQ3_M | 3.52GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q3_K.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q3_K.gguf) | Q3_K | 3.74GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q3_K_M.gguf) | Q3_K_M | 3.74GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q3_K_L.gguf) | Q3_K_L | 4.03GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.IQ4_XS.gguf) | IQ4_XS | 4.18GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q4_0.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q4_0.gguf) | Q4_0 | 4.34GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.IQ4_NL.gguf) | IQ4_NL | 4.38GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q4_K_S.gguf) | Q4_K_S | 4.37GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q4_K.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q4_K.gguf) | Q4_K | 4.58GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q4_K_M.gguf) | Q4_K_M | 4.58GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q4_1.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q4_1.gguf) | Q4_1 | 4.78GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q5_0.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q5_0.gguf) | Q5_0 | 5.21GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q5_K_S.gguf) | Q5_K_S | 5.21GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q5_K.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q5_K.gguf) | Q5_K | 5.34GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q5_K_M.gguf) | Q5_K_M | 5.34GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q5_1.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q5_1.gguf) | Q5_1 | 5.65GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q6_K.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q6_K.gguf) | Q6_K | 6.14GB |\n| [llama3-8b-cpt-sea-lionv2.1-instruct.Q8_0.gguf](https://huggingface.co/RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf/blob/main/llama3-8b-cpt-sea-lionv2.1-instruct.Q8_0.gguf) | Q8_0 | 7.95GB |\n\n\n\n\nOriginal model description:\n---\nlanguage:\n- en\n- id\n- ta\n- th\n- vi\nlicense: llama3\n---\n# Llama3 8B CPT SEA-Lionv2.1 Instruct\n\nSEA-LION is a collection of Large Language Models (LLMs) which has been pretrained and instruct-tuned for the Southeast Asia (SEA) region.\n\nLlama3 8B CPT SEA-Lionv2.1 Instruct is a multilingual model which has been fine-tuned with around **100,000 English instruction-completion pairs** alongside a smaller pool of around **50,000 instruction-completion pairs** from other ASEAN languages, such as Indonesian, Thai and Vietnamese.\nThese instructions have been carefully curated and rewritten to ensure the model was trained on truly open, commercially permissive and high quality datasets.\n\nLlama3 8B CPT SEA-Lionv2.1 Instruct has undergone additional supervised fine-tuning and alignment compared to the now deprecated Llama3 8B CPT SEA-Lionv2 Instruct. These improvements have increased the model's capabilities in chat interactions and its ability to follow instructions accurately.\n\nSEA-LION stands for _Southeast Asian Languages In One Network_.\n\n- **Developed by:** Products Pillar, AI Singapore\n- **Funded by:** Singapore NRF\n- **Model type:** Decoder\n- **Languages:** English, Indonesian, Thai, Vietnamese, Tamil\n- **License:** [Llama3 Community License](https://huggingface.co/meta-llama/Meta-Llama-3-8B/blob/main/LICENSE)\n\n## Model Details\n\n### Model Description\nWe performed instruction tuning in English and also in ASEAN languages such as Indonesian, Thai and Vietnamese on our [continued pre-trained Llama3 CPT 8B SEA-Lionv2](https://huggingface.co/aisingapore/llama3-8b-cpt-SEA-Lionv2-base), a decoder model using the Llama3 architecture, to create Llama3 8B SEA-Lionv2.1 Instruct.\n\nThe model has a context length of 8192.\n\n### Benchmark Performance\nWe evaluated Llama3 8B SEA-Lionv2.1 Instruct on both general language capabilities and instruction-following capabilities.\n\n#### General Language Capabilities\nFor the evaluation of general language capabilities, we employed the [BHASA evaluation benchmark](https://arxiv.org/abs/2309.06085v2) across a variety of tasks.\nThese tasks include Question Answering (QA), Sentiment Analysis (Sentiment), Toxicity Detection (Toxicity), Translation in both directions (Eng>Lang & Lang>Eng), Abstractive Summarization (Summ), Causal Reasoning (Causal) and Natural Language Inference (NLI).\n\nNote: BHASA is implemented following a strict answer format, and only spaces and punctuations are cleaned. For tasks where options are provided, the answer should only include one of the pre-defined options, nothing else. If the model continues to generate more tokens (e.g. to explain its answer), it will be considered to be a wrong response. For the F1 score metric (as used in Sentiment Analysis and Toxicity Detection), all answers that do not fall under the pre-defined labels will be treated as a separate label (to mark it as a wrong answer) and included in the calculations so that the model is penalized for not generating one of the pre-defined labels.\n\nThe evaluation was done zero-shot with native prompts and only a sample of 100-1000 instances for each dataset was used as per the setting described in the paper.\n\n\n#### Instruction-following Capabilities\nSince LLama3 8B SEA-Lionv2.1 is an instruction-following model, we also evaluated it on instruction-following capabilities with two datasets, [IFEval](https://arxiv.org/abs/2311.07911) and [MT-Bench](https://arxiv.org/abs/2306.05685).\n\nAs these two datasets were originally in English, the linguists and native speakers in the team worked together to filter, localize and translate the datasets into the respective target languages to ensure that the examples remained reasonable, meaningful and natural.\n\n**IFEval**\n\nIFEval evaluates a model's ability to adhere to constraints provided in the prompt, for example beginning a response with a specific word/phrase or answering with a certain number of sections. The metric used is accuracy normalized by language (if the model performs the task correctly but responds in the wrong language, it is judged to have failed the task).\n\n\n**MT-Bench**\n\nMT-Bench evaluates a model's ability to engage in multi-turn (2 turns) conversations and respond in ways that align with human needs. We use `gpt-4-1106-preview` as the judge model and compare against `gpt-3.5-turbo-0125` as the baseline model. The metric used is the weighted win rate against the baseline model (i.e. average win rate across each category (Math, Reasoning, STEM, Humanities, Roleplay, Writing, Extraction)). A tie is given a score of 0.5.\n\n\nFor more details on Llama3 8B CPT SEA-Lionv2.1 Instruct benchmark performance, please refer to the SEA HELM leaderboard, https://leaderboard.sea-lion.ai/\n\n\n### Usage\nSEA-LION can be run using the 🤗 Transformers library \n```python\n# Please use transformers==4.43.2\n\nimport transformers\nimport torch\n\nmodel_id = \"aisingapore/llama3-8b-cpt-SEA-Lionv2.1-instruct\"\n\npipeline = transformers.pipeline(\n    \"text-generation\",\n    model=model_id,\n    model_kwargs={\"torch_dtype\": torch.bfloat16},\n    device_map=\"auto\",\n)\nmessages = [\n    {\"role\": \"user\", \"content\": \"Apa sentimen dari kalimat berikut ini?\\nKalimat: Buku ini sangat membosankan.\\nJawaban: \"},\n]\n\noutputs = pipeline(\n    messages,\n    max_new_tokens=256,\n)\nprint(outputs[0][\"generated_text\"][-1])\n```\n\n### Accessing Older Revisions\nHuggingface provides support for the revision parameter, allowing users to access specific versions of models. This can be used to retrieve the original llama3-8b-cpt-SEA-Lionv2-instruct model with the tag \"v2.0.0\".\n```python\n# Please use transformers==4.43.2\n\nimport transformers\nimport torch\n\nmodel_id = \"aisingapore/llama3-8b-cpt-SEA-Lionv2.1-instruct\"\n\npipeline = transformers.pipeline(\n    \"text-generation\",\n    model=model_id,\n    model_kwargs={\"torch_dtype\": torch.bfloat16, \"revision\": \"v2.0.0\"},\n    device_map=\"auto\",\n)\nmessages = [\n    {\"role\": \"user\", \"content\": \"Apa sentimen dari kalimat berikut ini?\\nKalimat: Buku ini sangat membosankan.\\nJawaban: \"},\n]\n\noutputs = pipeline(\n    messages,\n    max_new_tokens=256,\n)\nprint(outputs[0][\"generated_text\"][-1])\n```\n\n### Caveats\nIt is important for users to be aware that our model exhibits certain limitations that warrant consideration. Like many LLMs, the model can hallucinate and occasionally generates irrelevant content, introducing fictional elements that are not grounded in the provided context. Users should also exercise caution in interpreting and validating the model's responses due to the potential inconsistencies in its reasoning.\n\n## Limitations\n### Safety\n\nCurrent SEA-LION models, including this commercially permissive release, have not been aligned for safety. Developers and users should perform their own safety fine-tuning and related security measures. In no event shall the authors be held liable for any claim, damages, or other liability arising from the use of the released weights and codes.\n\n## Technical Specifications\n### Fine-Tuning Details\nThe Llama3 8B CPT SEA-Lionv2.1 Instruct was fine-tuned using 8x A100-40GB using parameter efficient fine tuning in the form of LoRA.\n\n## Data\nLlama3 8B CPT SEA-Lionv2.1 Instruct was trained on a wide range of instructions that were manually and stringently verified by our team. A large portion of the effort was dedicated to ensuring that each instruction-completion pair that the model sees is of high quality and any errors were corrected and rewritten by native speakers or else dropped from our mix.\n\nIn addition, special care was taken to ensure that the datasets used had commercially permissive licenses through verification with the original data source. \n\nLink to dataset: _coming soon_\n\n## Call for Contributions\nWe encourage researchers, developers, and language enthusiasts to actively contribute to the enhancement and expansion of SEA-LION. Contributions can involve identifying and reporting bugs, sharing pre-training, instruction, and preference data, improving documentation usability, proposing and implementing new model evaluation tasks and metrics, or training versions of the model in additional Southeast Asian languages. Join us in shaping the future of SEA-LION by sharing your expertise and insights to make these models more accessible, accurate, and versatile. Please check out our GitHub for further information on the call for contributions.\n\n## The Team\n\nChoa Esther<br> \nCheng Nicholas<br> \nHuang Yuli<br> \nLau Wayne<br> \nLee Chwan Ren<br> \nLeong Wai Yi<br> \nLeong Wei Qi<br> \nLi Yier<br> \nLiu Bing Jie Darius<br> \nLovenia Holy<br> \nMontalan Jann Railey<br> \nNg Boon Cheong Raymond<br> \nNgui Jian Gang<br> \nNguyen Thanh Ngan<br> \nOng Brandon<br> \nOng Tat-Wee David<br> \nOng Zhi Hao<br> \nRengarajan Hamsawardhini<br> \nSiow Bryan<br> \nSusanto Yosephine<br> \nTai Ngee Chia<br> \nTan Choon Meng<br> \nTeo Eng Sipp Leslie<br> \nTeo Wei Yi<br> \nTjhi William<br> \nTeng Walter<br> \nYeo Yeow Tong<br> \nYong Xianbin<br>\n\n## Acknowledgements\n\n[AI Singapore](​​https://aisingapore.org/) is a national programme supported by the National Research Foundation, Singapore and hosted by the National University of Singapore. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of the National Research Foundation or the National University of Singapore. \n\n## Contact\n\nFor more info, please contact us using this [SEA-LION Inquiry Form](https://forms.gle/sLCUVb95wmGf43hi6)\n\n[Link to SEA-LION's GitHub repository](https://github.com/aisingapore/sealion)\n\n## Disclaimer\n\nThis is the repository for the commercial instruction-tuned model.\nThe model has _not_ been aligned for safety.\nDevelopers and users should perform their own safety fine-tuning and related security measures.\nIn no event shall the authors be held liable for any claims, damages, or other liabilities arising from the use of the released weights and codes.\n\n",
    "related_quantizations": []
  },
  "tags": [
    "gguf",
    "arxiv:2309.06085",
    "arxiv:2311.07911",
    "arxiv:2306.05685",
    "endpoints_compatible",
    "region:us",
    "conversational"
  ],
  "likes": 0,
  "downloads": 232,
  "gated": false,
  "private": false,
  "last_modified": "2024-08-23T17:25:06.000Z",
  "created_at": "2024-08-23T14:18:58.000Z",
  "pipeline_tag": "",
  "library_name": ""
}

{
  "_id": "66c89a5271ee4daba2e3db37",
  "id": "RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf",
  "modelId": "RichardErkhov/aisingapore_-_llama3-8b-cpt-sea-lionv2.1-instruct-gguf",
  "sha": "943676a9f05c22b6da92fa1c6b8c046aa9ac7401",
  "createdAt": "2024-08-23T14:18:58.000Z",
  "lastModified": "2024-08-23T17:25:06.000Z",
  "author": "RichardErkhov",
  "downloads": 232,
  "likes": 0,
  "gated": false,
  "private": false,
  "pipeline_tag": "",
  "library_name": "",
  "siblings_count": 24
}