richarderkhov/swap-uniba_-_llamantino-2-chat-13b-hf-ultrachat-ita-gguf overview

{
  "metadata": {},
  "card_data": {
    "frontmatter": {},
    "hero_image_url": "https://i.ibb.co/6mHSRm3/llamantino53.jpg",
    "summary": "*Last Update: 08/01/2024* *Example of Use*: Colab Notebook",
    "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\nLLaMAntino-2-chat-13b-hf-UltraChat-ITA - GGUF\n- Model creator: https://huggingface.co/swap-uniba/\n- Original model: https://huggingface.co/swap-uniba/LLaMAntino-2-chat-13b-hf-UltraChat-ITA/\n\n\n| Name | Quant method | Size |\n| ---- | ---- | ---- |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q2_K.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q2_K.gguf) | Q2_K | 4.52GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.IQ3_XS.gguf) | IQ3_XS | 4.99GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.IQ3_S.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.IQ3_S.gguf) | IQ3_S | 5.27GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q3_K_S.gguf) | Q3_K_S | 5.27GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.IQ3_M.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.IQ3_M.gguf) | IQ3_M | 5.57GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q3_K.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q3_K.gguf) | Q3_K | 5.9GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q3_K_M.gguf) | Q3_K_M | 5.9GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q3_K_L.gguf) | Q3_K_L | 6.45GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.IQ4_XS.gguf) | IQ4_XS | 6.54GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q4_0.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q4_0.gguf) | Q4_0 | 6.86GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.IQ4_NL.gguf) | IQ4_NL | 6.9GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q4_K_S.gguf) | Q4_K_S | 6.91GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q4_K.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q4_K.gguf) | Q4_K | 7.33GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q4_K_M.gguf) | Q4_K_M | 7.33GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q4_1.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q4_1.gguf) | Q4_1 | 7.61GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q5_0.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q5_0.gguf) | Q5_0 | 8.36GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q5_K_S.gguf) | Q5_K_S | 8.36GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q5_K.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q5_K.gguf) | Q5_K | 8.6GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q5_K_M.gguf) | Q5_K_M | 8.6GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q5_1.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q5_1.gguf) | Q5_1 | 9.1GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q6_K.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q6_K.gguf) | Q6_K | 9.95GB |\n| [LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q8_0.gguf](https://huggingface.co/RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf/blob/main/LLaMAntino-2-chat-13b-hf-UltraChat-ITA.Q8_0.gguf) | Q8_0 | 12.88GB |\n\n\n\n\nOriginal model description:\n---\nlicense: llama2\nlanguage:\n- it\ntags:\n- text-generation-inference\n---\n<img src=\"https://i.ibb.co/6mHSRm3/llamantino53.jpg\" alt=\"llamantino53\" border=\"0\" width=\"200px\">\n\n# Model Card for LLaMAntino-2-chat-13b-UltraChat-ITA\n*Last Update: 08/01/2024*<br>\n*Example of Use*: [Colab Notebook](https://colab.research.google.com/drive/1xUite70ANLQp8NwQE93jlI3epj_cpua7?usp=sharing)\n<hr>\n\n## Model description\n\n<!-- Provide a quick summary of what the model is/does. -->\n\n**LLaMAntino-2-chat-13b-UltraChat** is a *Large Language Model (LLM)* that is an instruction-tuned version of **LLaMAntino-2-chat-13b** (an italian-adapted **LLaMA 2 chat**). \nThis model aims to provide Italian NLP researchers with an improved model for italian dialogue use cases.\n\nThe model was trained using *QLora* and using as training data [UltraChat](https://github.com/thunlp/ultrachat) translated to the italian language using [Argos Translate](https://pypi.org/project/argostranslate/1.4.0/). \nIf you are interested in more details regarding the training procedure, you can find the code we used at the following link:\n- **Repository:** https://github.com/swapUniba/LLaMAntino\n\n**NOTICE**: the code has not been released yet, we apologize for the delay, it will be available asap!\n\n- **Developed by:** Pierpaolo Basile, Elio Musacchio, Marco Polignano, Lucia Siciliani, Giuseppe Fiameni, Giovanni Semeraro\n- **Funded by:** PNRR project FAIR - Future AI Research\n- **Compute infrastructure:** [Leonardo](https://www.hpc.cineca.it/systems/hardware/leonardo/) supercomputer\n- **Model type:** LLaMA-2-chat\n- **Language(s) (NLP):** Italian\n- **License:** Llama 2 Community License \n- **Finetuned from model:** [swap-uniba/LLaMAntino-2-chat-13b-hf-ITA](https://huggingface.co/swap-uniba/LLaMAntino-2-chat-13b-hf-ITA)\n\n## Prompt Format\n\nThis prompt format based on the [LLaMA 2 prompt template](https://gpus.llm-utils.org/llama-2-prompt-template/) adapted to the italian language was used:\n\n```python\n\" [INST]<<SYS>>\\n\" \\\n\"Sei un assistente disponibile, rispettoso e onesto. \" \\\n\"Rispondi sempre nel modo piu' utile possibile, pur essendo sicuro. \" \\\n\"Le risposte non devono includere contenuti dannosi, non etici, razzisti, sessisti, tossici, pericolosi o illegali. \" \\\n\"Assicurati che le tue risposte siano socialmente imparziali e positive. \" \\\n\"Se una domanda non ha senso o non e' coerente con i fatti, spiegane il motivo invece di rispondere in modo non corretto. \" \\\n\"Se non conosci la risposta a una domanda, non condividere informazioni false.\\n\" \\\n\"<</SYS>>\\n\\n\" \\\nf\"{user_msg_1}[/INST] {model_answer_1} </s> <s> [INST]{user_msg_2}[/INST] {model_answer_2} </s> ... <s> [INST]{user_msg_N}[/INST] {model_answer_N} </s>\"\n```\n\nWe recommend using the same prompt in inference to obtain the best results!\n\n## How to Get Started with the Model\n\nBelow you can find an example of model usage:\n\n```python\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\n\nmodel_id = \"swap-uniba/LLaMAntino-2-chat-13b-hf-UltraChat-ITA\"\n\ntokenizer = AutoTokenizer.from_pretrained(model_id)\nmodel = AutoModelForCausalLM.from_pretrained(model_id)\n\nuser_msg = \"Ciao! Come stai?\"\n\nprompt = \" [INST]<<SYS>>\\n\" \\\n         \"Sei un assistente disponibile, rispettoso e onesto. \" \\\n         \"Rispondi sempre nel modo piu' utile possibile, pur essendo sicuro. \" \\\n         \"Le risposte non devono includere contenuti dannosi, non etici, razzisti, sessisti, tossici, pericolosi o illegali. \" \\\n         \"Assicurati che le tue risposte siano socialmente imparziali e positive. \" \\\n         \"Se una domanda non ha senso o non e' coerente con i fatti, spiegane il motivo invece di rispondere in modo non corretto. \" \\\n         \"Se non conosci la risposta a una domanda, non condividere informazioni false.\\n\" \\\n         \"<</SYS>>\\n\\n\" \\\n         f\"{user_msg}[/INST]\"\n\npipe = transformers.pipeline(\n    model=model,\n    tokenizer=tokenizer,\n    return_full_text=False, # langchain expects the full text\n    task='text-generation',\n    max_new_tokens=512, # max number of tokens to generate in the output\n    temperature=0.8  #temperature for more or less creative answers\n)\n\n# Method 1\nsequences = pipe(text)\nfor seq in sequences:\n    print(f\"{seq['generated_text']}\")\n\n# Method 2\ninput_ids = tokenizer(prompt, return_tensors=\"pt\").input_ids\noutputs = model.generate(input_ids=input_ids, max_length=512)\nprint(tokenizer.batch_decode(outputs.detach().cpu().numpy()[:, input_ids.shape[1]:], skip_special_tokens=True)[0])\n```\n\nIf you are facing issues when loading the model, you can try to load it **Quantized**:\n\n```python\nmodel = AutoModelForCausalLM.from_pretrained(model_id, load_in_8bit=True)\n```\n\n*Note*:\n1) The model loading strategy above requires the [*bitsandbytes*](https://pypi.org/project/bitsandbytes/) and [*accelerate*](https://pypi.org/project/accelerate/) libraries\n2) The Tokenizer, by default, adds at the beginning of the prompt the '\\<BOS\\>' token. If that is not the case, add as a starting token the *\\<s\\>* string.\n   \n## Evaluation\n\n<!-- This section describes the evaluation protocols and provides the results. -->\n\n*Coming soon*!\n\n## Citation\n\n<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->\n\nIf you use this model in your research, please cite the following:\n\n```bibtex\n@misc{basile2023llamantino,\n      title={LLaMAntino: LLaMA 2 Models for Effective Text Generation in Italian Language}, \n      author={Pierpaolo Basile and Elio Musacchio and Marco Polignano and Lucia Siciliani and Giuseppe Fiameni and Giovanni Semeraro},\n      year={2023},\n      eprint={2312.09993},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n```\n\n*Notice:* Llama 2 is licensed under the LLAMA 2 Community License, Copyright © Meta Platforms, Inc. All Rights Reserved. [*License*](https://ai.meta.com/llama/license/)\n\n\n",
    "related_quantizations": []
  },
  "tags": [
    "gguf",
    "arxiv:2312.09993",
    "endpoints_compatible",
    "region:us"
  ],
  "likes": 0,
  "downloads": 702,
  "gated": false,
  "private": false,
  "last_modified": "2024-08-03T16:29:30.000Z",
  "created_at": "2024-08-03T11:02:31.000Z",
  "pipeline_tag": "",
  "library_name": ""
}

{
  "_id": "66ae0e476038fe024a620879",
  "id": "RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf",
  "modelId": "RichardErkhov/swap-uniba_-_LLaMAntino-2-chat-13b-hf-UltraChat-ITA-gguf",
  "sha": "6f060f91de3a81b3e12f02bbdde8d8eb4c21a090",
  "createdAt": "2024-08-03T11:02:31.000Z",
  "lastModified": "2024-08-03T16:29:30.000Z",
  "author": "RichardErkhov",
  "downloads": 702,
  "likes": 0,
  "gated": false,
  "private": false,
  "pipeline_tag": "",
  "library_name": "",
  "siblings_count": 24
}