notebooks/llm-chatbot/llm-chatbot-generate-api.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "50e1d1d5-3bdd-4224-9f93-bf5d9a83f424",
   "metadata": {},
   "source": [
    "# Create an LLM-powered Chatbot using OpenVINO Generate API\n",
    "\n",
    "In the rapidly evolving world of artificial intelligence (AI), chatbots have emerged as powerful tools for businesses to enhance customer interactions and streamline operations. \n",
    "Large Language Models (LLMs) are artificial intelligence systems that can understand and generate human language. They use deep learning algorithms and massive amounts of data to learn the nuances of language and produce coherent and relevant responses.\n",
    "While a decent intent-based chatbot can answer basic, one-touch inquiries like order management, FAQs, and policy questions, LLM chatbots can tackle more complex, multi-touch questions. LLM enables chatbots to provide support in a conversational manner, similar to how humans do, through contextual memory. Leveraging the capabilities of Language Models, chatbots are becoming increasingly intelligent, capable of understanding and responding to human language with remarkable accuracy.\n",
    "\n",
    "Previously, we already discussed how to build an instruction-following pipeline using OpenVINO, please check out [this tutorial](../llm-question-answering/llm-question-answering.ipynb) for reference.\n",
    "In this tutorial, we consider how to use the power of OpenVINO for running Large Language Models for chat. We will use a pre-trained model from the [Hugging Face Transformers](https://huggingface.co/docs/transformers/index) library. The [Hugging Face Optimum Intel](https://huggingface.co/docs/optimum/intel/index) library converts the models to OpenVINO™ IR format. To simplify the user experience, we will use [OpenVINO Generate API](https://github.com/openvinotoolkit/openvino.genai) for generation pipeline.\n",
    "\n",
    "\n",
    "The tutorial consists of the following steps:\n",
    "\n",
    "- Install prerequisites\n",
    "- Download and convert the model from a public source using the [OpenVINO integration with Hugging Face Optimum](https://huggingface.co/blog/openvino).\n",
    "- Compress model weights to 4-bit or 8-bit data types using [NNCF](https://github.com/openvinotoolkit/nncf)\n",
    "- Create a chat inference pipeline with [OpenVINO Generate API](https://github.com/openvinotoolkit/openvino.genai/blob/master/src/README.md).\n",
    "- Run chat pipeline with [Gradio](https://www.gradio.app/).\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "### Installation Instructions\n",
    "\n",
    "This is a self-contained example that relies solely on its own code.\n",
    "\n",
    "We recommend  running the notebook in a virtual environment. You only need a Jupyter server to start.\n",
    "For details, please refer to [Installation Guide](https://github.com/openvinotoolkit/openvino_notebooks/blob/latest/README.md#-installation-guide).\n",
    "\n",
    "<img referrerpolicy=\"no-referrer-when-downgrade\" src=\"https://static.scarf.sh/a.png?x-pxid=5b5a4db0-7875-4bfb-bdbd-01698b5b1a77&file=notebooks/llm-chatbot/llm-chatbot-generate-api.ipynb\" />\n",
    "\n",
    "#### Table of contents:\n",
    "\n",
    "- [Prerequisites](#Prerequisites)\n",
    "- [Select model for inference](#Select-model-for-inference)\n",
    "- [Convert model using Optimum-CLI tool](#Convert-model-using-Optimum-CLI-tool)\n",
    "    - [Weights Compression using Optimum-CLI](#Weights-Compression-using-Optimum-CLI)\n",
    "- [Select device for inference](#Select-device-for-inference)\n",
    "- [Instantiate pipeline with OpenVINO Generate API](#Instantiate-pipeline-with-OpenVINO-Generate-API)\n",
    "- [Run Chatbot](#Run-Chatbot)\n",
    "    - [Advanced generation options](#Advanced-generation-options)\n",
    "\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "5df233b0-0369-4fff-9952-7957a90394a5",
   "metadata": {},
   "source": [
    "## Prerequisites\n",
    "[back to top ⬆️](#Table-of-contents:)\n",
    "\n",
    "Install required dependencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "563ecf9f-346b-4f14-85ef-c66ff0c95f65",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "os.environ[\"GIT_CLONE_PROTECTION_ACTIVE\"] = \"false\"\n",
    "\n",
    "%pip install -Uq pip\n",
    "%pip uninstall -q -y optimum optimum-intel\n",
    "%pip install -q -U \"openvino>=2024.3.0\" openvino-tokenizers[transformers] openvino-genai\n",
    "%pip install -q --extra-index-url https://download.pytorch.org/whl/cpu\\\n",
    "\"git+https://github.com/huggingface/optimum-intel.git\"\\\n",
    "\"git+https://github.com/openvinotoolkit/nncf.git\"\\\n",
    "\"torch>=2.1\"\\\n",
    "\"datasets\" \\\n",
    "\"accelerate\" \\\n",
    "\"gradio>=4.19\" \\\n",
    "\"transformers>=4.43.1\" \\\n",
    "\"huggingface-hub>=0.26.5\" \\\n",
    "\"onnx<=1.16.1; sys_platform=='win32'\" \"einops\" \"transformers_stream_generator\" \"tiktoken\" \"bitsandbytes\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "f39ca954-61d2-45c5-a7f9-7fce1acc277f",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import os\n",
    "from pathlib import Path\n",
    "import requests\n",
    "import shutil\n",
    "\n",
    "# fetch model configuration\n",
    "\n",
    "config_shared_path = Path(\"../../utils/llm_config.py\")\n",
    "config_dst_path = Path(\"llm_config.py\")\n",
    "\n",
    "if not config_dst_path.exists():\n",
    "    if config_shared_path.exists():\n",
    "        try:\n",
    "            os.symlink(config_shared_path, config_dst_path)\n",
    "        except Exception:\n",
    "            shutil.copy(config_shared_path, config_dst_path)\n",
    "    else:\n",
    "        r = requests.get(url=\"https://raw.githubusercontent.com/openvinotoolkit/openvino_notebooks/latest/utils/llm_config.py\")\n",
    "        with open(\"llm_config.py\", \"w\", encoding=\"utf-8\") as f:\n",
    "            f.write(r.text)\n",
    "elif not os.path.islink(config_dst_path):\n",
    "    print(\"LLM config will be updated\")\n",
    "    if config_shared_path.exists():\n",
    "        shutil.copy(config_shared_path, config_dst_path)\n",
    "    else:\n",
    "        r = requests.get(url=\"https://raw.githubusercontent.com/openvinotoolkit/openvino_notebooks/latest/utils/llm_config.py\")\n",
    "        with open(\"llm_config.py\", \"w\", encoding=\"utf-8\") as f:\n",
    "            f.write(r.text)\n",
    "\n",
    "if not Path(\"notebook_utils.py\").exists():\n",
    "    r = requests.get(url=\"https://raw.githubusercontent.com/openvinotoolkit/openvino_notebooks/latest/utils/notebook_utils.py\")\n",
    "    open(\"notebook_utils.py\", \"w\").write(r.text)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "6d62f9f4-5434-4550-b372-c86b5a5089d5",
   "metadata": {},
   "source": [
    "## Select device for inference\n",
    "[back to top ⬆️](#Table-of-contents:)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "07cddc37",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "650181049bdb4ffcb84afc9aee554261",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Dropdown(description='Device:', options=('CPU', 'AUTO'), value='CPU')"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from notebook_utils import device_widget\n",
    "\n",
    "device = device_widget(default=\"CPU\")\n",
    "\n",
    "device"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "81983176-e571-4652-ba21-4bd608c35146",
   "metadata": {},
   "source": [
    "## Select model for inference\n",
    "[back to top ⬆️](#Table-of-contents:)\n",
    "\n",
    "The tutorial supports different models, you can select one from the provided options to compare the quality of open source LLM solutions.\n",
    "Model conversion and optimization is time- and memory-consuming process.  For your convenience, we provide a [collection](https://huggingface.co/collections/OpenVINO/llm-6687aaa2abca3bbcec71a9bd) of optimized models on HuggingFace hub. You can skip the model conversion step by selecting one of the available on HuggingFace hub model. If you want to reproduce optimization process locally, please unset **Use preconverted models** checkbox.\n",
    "\n",
    ">**Note**: conversion of some models can require additional actions from user side and at least 64GB RAM for conversion.\n",
    "\n",
    "[Weight compression](https://docs.openvino.ai/2024/openvino-workflow/model-optimization-guide/weight-compression.html) is a technique for enhancing the efficiency of models, especially those with large memory requirements. This method reduces the model’s memory footprint, a crucial factor for Large Language Models (LLMs). We provide several options for model weight compression:\n",
    "\n",
    "* **FP16** reducing model binary size on disk using `save_model` with enabled compression weights to FP16 precision. This approach is available in OpenVINO from scratch and is the default behavior.\n",
    "* **INT8** is an 8-bit weight-only quantization provided by [NNCF](https://github.com/openvinotoolkit/nncf): This method compresses weights to an 8-bit integer data type, which balances model size reduction and accuracy, making it a versatile option for a broad range of applications.\n",
    "* **INT4** is an 4-bit weight-only quantization provided by [NNCF](https://github.com/openvinotoolkit/nncf). involves quantizing weights to an unsigned 4-bit integer symmetrically around a fixed zero point of eight (i.e., the midpoint between zero and 15). in case of **symmetric quantization** or asymmetrically with a non-fixed zero point, in case of **asymmetric quantization** respectively. Compared to INT8 compression, INT4 compression improves performance even more, but introduces a minor drop in prediction quality. INT4 it ideal for situations where speed is prioritized over an acceptable trade-off against accuracy.\n",
    "* **INT4 AWQ** is an 4-bit activation-aware weight quantization. [Activation-aware Weight Quantization](https://arxiv.org/abs/2306.00978) (AWQ) is an algorithm that tunes model weights for more accurate INT4 compression. It slightly improves generation quality of compressed LLMs, but requires significant additional time for tuning weights on a calibration dataset. We will use `wikitext-2-raw-v1/train` subset of the [Wikitext](https://huggingface.co/datasets/Salesforce/wikitext) dataset for calibration.\n",
    "* **INT4 NPU-friendly** is an 4-bit channel-wise quantization. This approach is [recommended](https://docs.openvino.ai/2024/learn-openvino/llm_inference_guide/genai-guide-npu.html) for LLM inference using NPU.\n",
    "\n",
    "<details>\n",
    "  <summary><b>Click here to see available models options</b></summary>\n",
    "\n",
    "* **tiny-llama-1b-chat** - This is the chat model finetuned on top of [TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T](https://huggingface.co/TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T). The TinyLlama project aims to pretrain a 1.1B Llama model on 3 trillion tokens with the adoption of the same architecture and tokenizer as Llama 2. This means TinyLlama can be plugged and played in many open-source projects built upon Llama. Besides, TinyLlama is compact with only 1.1B parameters. This compactness allows it to cater to a multitude of applications demanding a restricted computation and memory footprint. More details about model can be found in [model card](https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0)\n",
    "*  **minicpm-2b-dpo** - MiniCPM is an End-Size LLM developed by ModelBest Inc. and TsinghuaNLP, with only 2.4B parameters excluding embeddings. After Direct Preference Optimization (DPO) fine-tuning, MiniCPM outperforms many popular 7b, 13b and 70b models. More details can be found in [model_card](https://huggingface.co/openbmb/MiniCPM-2B-dpo-fp16).\n",
    "*  **minicpm3-4b** - MiniCPM3-4B is the 3rd generation of MiniCPM series. The overall performance of MiniCPM3-4B surpasses Phi-3.5-mini-Instruct, being comparable with many recent 7B~9B models.Compared to previous generations, MiniCPM3-4B has a more powerful and versatile skill set to enable more general usage. More details can be found in [model card](https://huggingface.co/openbmb/MiniCPM3-4B).\n",
    "* **llama-3.2-1B-instruct** - 1B parameters model from LLama3.2 collection of instruction-tuned multilingual models. Llama 3.2 instruction-tuned text only models are optimized for multilingual dialogue use cases, including agentic retrieval and summarization tasks. They outperform many of the available open source and closed chat models on common industry benchmarks. More details can be found in [model card](https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct)\n",
    ">**Note**: run model with demo, you will need to accept license agreement. \n",
    ">You must be a registered user in 🤗 Hugging Face Hub. Please visit [HuggingFace model card](https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct), carefully read terms of usage and click accept button.  You will need to use an access token for the code below to run. For more information on access tokens, refer to [this section of the documentation](https://huggingface.co/docs/hub/security-tokens).\n",
    ">You can login on Hugging Face Hub in notebook environment, using following code:\n",
    " \n",
    "```python\n",
    "    # login to huggingfacehub to get access to pretrained model \n",
    "\n",
    "\n",
    "    from huggingface_hub import notebook_login, whoami\n",
    "\n",
    "    try:\n",
    "        whoami()\n",
    "        print('Authorization token already provided')\n",
    "    except OSError:\n",
    "        notebook_login()\n",
    "```\n",
    "* **llama-3.2-3B-instruct** - 3B parameters model from LLama3.2 collection of instruction-tuned multilingual models. Llama 3.2 instruction-tuned text only models are optimized for multilingual dialogue use cases, including agentic retrieval and summarization tasks. They outperform many of the available open source and closed chat models on common industry benchmarks. More details can be found in [model card](https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct)\n",
    ">**Note**: run model with demo, you will need to accept license agreement. \n",
    ">You must be a registered user in 🤗 Hugging Face Hub. Please visit [HuggingFace model card](https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct), carefully read terms of usage and click accept button.  You will need to use an access token for the code below to run. For more information on access tokens, refer to [this section of the documentation](https://huggingface.co/docs/hub/security-tokens).\n",
    ">You can login on Hugging Face Hub in notebook environment, using following code:\n",
    " \n",
    "```python\n",
    "    # login to huggingfacehub to get access to pretrained model \n",
    "\n",
    "\n",
    "    from huggingface_hub import notebook_login, whoami\n",
    "\n",
    "    try:\n",
    "        whoami()\n",
    "        print('Authorization token already provided')\n",
    "    except OSError:\n",
    "        notebook_login()\n",
    "```\n",
    "\n",
    "*  **gemma-2b-it** - Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models. They are text-to-text, decoder-only large language models, available in English, with open weights, pre-trained variants, and instruction-tuned variants. Gemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning. This model is instruction-tuned version of 2B parameters model. More details about model can be found in [model card](https://huggingface.co/google/gemma-2b-it).\n",
    ">**Note**: run model with demo, you will need to accept license agreement. \n",
    ">You must be a registered user in 🤗 Hugging Face Hub. Please visit [HuggingFace model card](https://huggingface.co/google/gemma-2b-it), carefully read terms of usage and click accept button.  You will need to use an access token for the code below to run. For more information on access tokens, refer to [this section of the documentation](https://huggingface.co/docs/hub/security-tokens).\n",
    ">You can login on Hugging Face Hub in notebook environment, using following code:\n",
    " \n",
    "```python\n",
    "    # login to huggingfacehub to get access to pretrained model \n",
    "\n",
    "\n",
    "    from huggingface_hub import notebook_login, whoami\n",
    "\n",
    "    try:\n",
    "        whoami()\n",
    "        print('Authorization token already provided')\n",
    "    except OSError:\n",
    "        notebook_login()\n",
    "```\n",
    "*  **gemma-2-2b-it** - Gemma2 is the second generation of a Gemma family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models. They are text-to-text, decoder-only large language models, available in English, with open weights, pre-trained variants, and instruction-tuned variants. Gemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning. This model is instruction-tuned version of 2B parameters model. More details about model can be found in [model card](https://huggingface.co/google/gemma-2-2b-it).\n",
    ">**Note**: run model with demo, you will need to accept license agreement. \n",
    ">You must be a registered user in 🤗 Hugging Face Hub. Please visit [HuggingFace model card](https://huggingface.co/google/gemma-2-2b-it), carefully read terms of usage and click accept button.  You will need to use an access token for the code below to run. For more information on access tokens, refer to [this section of the documentation](https://huggingface.co/docs/hub/security-tokens).\n",
    ">You can login on Hugging Face Hub in notebook environment, using following code:\n",
    " \n",
    "```python\n",
    "    # login to huggingfacehub to get access to pretrained model \n",
    "\n",
    "\n",
    "    from huggingface_hub import notebook_login, whoami\n",
    "\n",
    "    try:\n",
    "        whoami()\n",
    "        print('Authorization token already provided')\n",
    "    except OSError:\n",
    "        notebook_login()\n",
    "```\n",
    "* **phi-3-mini-instruct** - The Phi-3-Mini is a 3.8B parameters, lightweight, state-of-the-art open model trained with the Phi-3 datasets that includes both synthetic data and the filtered publicly available websites data with a focus on high-quality and reasoning dense properties. More details about model can be found in [model card](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct), [Microsoft blog](https://aka.ms/phi3blog-april) and [technical report](https://aka.ms/phi3-tech-report).\n",
    "* **phi-3.5-mini-instruct** - Phi-3.5-mini is a lightweight, state-of-the-art open model built upon datasets used for Phi-3 - synthetic data and filtered publicly available websites - with a focus on very high-quality, reasoning dense data. The model belongs to the Phi-3 model family and supports 128K token context length. The model underwent a rigorous enhancement process, incorporating both supervised fine-tuning, proximal policy optimization, and direct preference optimization to ensure precise instruction adherence and robust safety measures. More details about model can be found in [model card](https://huggingface.co/microsoft/Phi-3.5-mini-instruct), [Microsoft blog](https://aka.ms/phi3.5-techblog) and [technical report](https://arxiv.org/abs/2404.14219).\n",
    "* **red-pajama-3b-chat** - A 2.8B parameter pre-trained language model based on GPT-NEOX architecture. It was developed by Together Computer and leaders from the open-source AI community. The model is fine-tuned on OASST1 and Dolly2 datasets to enhance chatting ability. More details about model can be found in [HuggingFace model card](https://huggingface.co/togethercomputer/RedPajama-INCITE-Chat-3B-v1).\n",
    "*  **gemma-7b-it** - Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models. They are text-to-text, decoder-only large language models, available in English, with open weights, pre-trained variants, and instruction-tuned variants. Gemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning. This model is instruction-tuned version of 7B parameters model. More details about model can be found in [model card](https://huggingface.co/google/gemma-7b-it).\n",
    ">**Note**: run model with demo, you will need to accept license agreement. \n",
    ">You must be a registered user in 🤗 Hugging Face Hub. Please visit [HuggingFace model card](https://huggingface.co/google/gemma-7b-it), carefully read terms of usage and click accept button.  You will need to use an access token for the code below to run. For more information on access tokens, refer to [this section of the documentation](https://huggingface.co/docs/hub/security-tokens).\n",
    ">You can login on Hugging Face Hub in notebook environment, using following code:\n",
    " \n",
    "```python\n",
    "    # login to huggingfacehub to get access to pretrained model \n",
    "\n",
    "    from huggingface_hub import notebook_login, whoami\n",
    "\n",
    "    try:\n",
    "        whoami()\n",
    "        print('Authorization token already provided')\n",
    "    except OSError:\n",
    "        notebook_login()\n",
    "```\n",
    "*  **gemma-2-9b-it** - Gemma2 is the second generation of a Gemma family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models. They are text-to-text, decoder-only large language models, available in English, with open weights, pre-trained variants, and instruction-tuned variants. Gemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning. This model is instruction-tuned version of 9B parameters model. More details about model can be found in [model card](https://huggingface.co/google/gemma-2-9b-it).\n",
    ">**Note**: run model with demo, you will need to accept license agreement. \n",
    ">You must be a registered user in 🤗 Hugging Face Hub. Please visit [HuggingFace model card](https://huggingface.co/google/gemma-2-2b-it), carefully read terms of usage and click accept button.  You will need to use an access token for the code below to run. For more information on access tokens, refer to [this section of the documentation](https://huggingface.co/docs/hub/security-tokens).\n",
    ">You can login on Hugging Face Hub in notebook environment, using following code:\n",
    " \n",
    "```python\n",
    "    # login to huggingfacehub to get access to pretrained model \n",
    "\n",
    "\n",
    "    from huggingface_hub import notebook_login, whoami\n",
    "\n",
    "    try:\n",
    "        whoami()\n",
    "        print('Authorization token already provided')\n",
    "    except OSError:\n",
    "        notebook_login()\n",
    "```\n",
    "\n",
    "* **llama-2-7b-chat** - LLama 2 is the second generation of LLama models developed by Meta. Llama 2 is a collection of pre-trained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. llama-2-7b-chat is 7 billions parameters version of LLama 2 finetuned and optimized for dialogue use case. More details about model can be found in the [paper](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/), [repository](https://github.com/facebookresearch/llama) and [HuggingFace model card](https://huggingface.co/meta-llama/Llama-2-7b-chat-hf).\n",
    ">**Note**: run model with demo, you will need to accept license agreement. \n",
    ">You must be a registered user in 🤗 Hugging Face Hub. Please visit [HuggingFace model card](https://huggingface.co/meta-llama/Llama-2-7b-chat-hf), carefully read terms of usage and click accept button.  You will need to use an access token for the code below to run. For more information on access tokens, refer to [this section of the documentation](https://huggingface.co/docs/hub/security-tokens).\n",
    ">You can login on Hugging Face Hub in notebook environment, using following code:\n",
    " \n",
    "```python\n",
    "    # login to huggingfacehub to get access to pretrained model \n",
    "\n",
    "    from huggingface_hub import notebook_login, whoami\n",
    "\n",
    "    try:\n",
    "        whoami()\n",
    "        print('Authorization token already provided')\n",
    "    except OSError:\n",
    "        notebook_login()\n",
    "```\n",
    "* **llama-3-8b-instruct** - Llama 3 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety. The Llama 3 instruction tuned models are optimized for dialogue use cases and outperform many of the available open source chat models on common industry benchmarks. More details about model can be found in [Meta blog post](https://ai.meta.com/blog/meta-llama-3/), [model website](https://llama.meta.com/llama3) and [model card](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct).\n",
    ">**Note**: run model with demo, you will need to accept license agreement. \n",
    ">You must be a registered user in 🤗 Hugging Face Hub. Please visit [HuggingFace model card](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct), carefully read terms of usage and click accept button.  You will need to use an access token for the code below to run. For more information on access tokens, refer to [this section of the documentation](https://huggingface.co/docs/hub/security-tokens).\n",
    ">You can login on Hugging Face Hub in notebook environment, using following code:\n",
    " \n",
    "```python\n",
    "    # login to huggingfacehub to get access to pretrained model \n",
    "\n",
    "    from huggingface_hub import notebook_login, whoami\n",
    "\n",
    "    try:\n",
    "        whoami()\n",
    "        print('Authorization token already provided')\n",
    "    except OSError:\n",
    "        notebook_login()\n",
    "```\n",
    "* **llama-3.1-8b-instruct** - The Llama 3.1 instruction tuned text only models (8B, 70B, 405B) are optimized for multilingual dialogue use cases and outperform many of the available open source and closed chat models on common industry benchmarks. More details about model can be found in [Meta blog post](https://ai.meta.com/blog/meta-llama-3-1/), [model website](https://llama.meta.com) and [model card](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct).\n",
    ">**Note**: run model with demo, you will need to accept license agreement. \n",
    ">You must be a registered user in 🤗 Hugging Face Hub. Please visit [HuggingFace model card](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct), carefully read terms of usage and click accept button.  You will need to use an access token for the code below to run. For more information on access tokens, refer to [this section of the documentation](https://huggingface.co/docs/hub/security-tokens).\n",
    ">You can login on Hugging Face Hub in notebook environment, using following code:\n",
    " \n",
    "```python\n",
    "    # login to huggingfacehub to get access to pretrained model \n",
    "\n",
    "    from huggingface_hub import notebook_login, whoami\n",
    "\n",
    "    try:\n",
    "        whoami()\n",
    "        print('Authorization token already provided')\n",
    "    except OSError:\n",
    "        notebook_login()\n",
    "```\n",
    "* **qwen2.5-0.5b-instruct/qwen2.5-1.5b-instruct/qwen2.5-3b-instruct/qwen2.5-7b-instruct/qwen2.5-14b-instruct** - Qwen2.5 is the latest series of Qwen large language models. Comparing with Qwen2, Qwen2.5 series brings significant improvements in coding, mathematics and general knowledge skills. Additionally, it brings long-context and multiple languages support including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more. \n",
    "For more details, please refer to [model_card](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct), [blog](https://qwenlm.github.io/blog/qwen2.5/), [GitHub](https://github.com/QwenLM/Qwen2.5), and [Documentation](https://qwen.readthedocs.io/en/latest/).\n",
    "* **qwen-7b-chat** - Qwen-7B is the 7B-parameter version of the large language model series, Qwen (abbr. Tongyi Qianwen), proposed by Alibaba Cloud. Qwen-7B is a Transformer-based large language model, which is pretrained on a large volume of data, including web texts, books, codes, etc. For more details about Qwen, please refer to the [GitHub](https://github.com/QwenLM/Qwen) code repository.\n",
    "* **chatglm3-6b** - ChatGLM3-6B is the latest open-source model in the ChatGLM series. While retaining many excellent features such as smooth dialogue and low deployment threshold from the previous two generations, ChatGLM3-6B employs a more diverse training dataset, more sufficient training steps, and a more reasonable training strategy. ChatGLM3-6B adopts a newly designed [Prompt format](https://github.com/THUDM/ChatGLM3/blob/main/PROMPT_en.md), in addition to the normal multi-turn dialogue. You can find more details about model in the [model card](https://huggingface.co/THUDM/chatglm3-6b)\n",
    "* **mistral-7b** - The Mistral-7B-v0.1 Large Language Model (LLM) is a pretrained generative text model with 7 billion parameters. You can find more details about model in the [model card](https://huggingface.co/mistralai/Mistral-7B-v0.1), [paper](https://arxiv.org/abs/2310.06825) and [release blog post](https://mistral.ai/news/announcing-mistral-7b/).\n",
    "* **zephyr-7b-beta** - Zephyr is a series of language models that are trained to act as helpful assistants. Zephyr-7B-beta is the second model in the series, and is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) that was trained on on a mix of publicly available, synthetic datasets using [Direct Preference Optimization (DPO)](https://arxiv.org/abs/2305.18290). You can find more details about model in [technical report](https://arxiv.org/abs/2310.16944) and [HuggingFace model card](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta).\n",
    "* **neural-chat-7b-v3-1** - Mistral-7b model fine-tuned using Intel Gaudi. The model fine-tuned on the open source dataset [Open-Orca/SlimOrca](https://huggingface.co/datasets/Open-Orca/SlimOrca) and aligned with [Direct Preference Optimization (DPO) algorithm](https://arxiv.org/abs/2305.18290). More details can be found in [model card](https://huggingface.co/Intel/neural-chat-7b-v3-1) and [blog post](https://medium.com/@NeuralCompressor/the-practice-of-supervised-finetuning-and-direct-preference-optimization-on-habana-gaudi2-a1197d8a3cd3).\n",
    "* **notus-7b-v1** - Notus is a collection of fine-tuned models using [Direct Preference Optimization (DPO)](https://arxiv.org/abs/2305.18290). and related [RLHF](https://huggingface.co/blog/rlhf) techniques. This model is the first version, fine-tuned with DPO over zephyr-7b-sft. Following a data-first approach, the only difference between Notus-7B-v1 and Zephyr-7B-beta is the preference dataset used for dDPO. Proposed approach for dataset creation helps to effectively fine-tune Notus-7b that surpasses Zephyr-7B-beta and Claude 2 on [AlpacaEval](https://tatsu-lab.github.io/alpaca_eval/). More details about model can be found in [model card](https://huggingface.co/argilla/notus-7b-v1).\n",
    "* **youri-7b-chat** - Youri-7b-chat is a Llama2 based model. [Rinna Co., Ltd.](https://rinna.co.jp/) conducted further pre-training for the Llama2 model with a mixture of English and Japanese datasets to improve Japanese task capability. The model is publicly released on Hugging Face hub. You can find detailed information at the [rinna/youri-7b-chat project page](https://huggingface.co/rinna/youri-7b). \n",
    "* **baichuan2-7b-chat** - Baichuan 2 is the new generation of large-scale open-source language models launched by [Baichuan Intelligence inc](https://www.baichuan-ai.com/home). It is trained on a high-quality corpus with 2.6 trillion tokens and has achieved the best performance in authoritative Chinese and English benchmarks of the same size.\n",
    "* **internlm2-chat-1.8b** - InternLM2 is the second generation InternLM series. Compared to the previous generation model, it shows significant improvements in various capabilities, including reasoning, mathematics, and coding. More details about model can be found in [model repository](https://huggingface.co/internlm).\n",
    "* **glm-4-9b-chat** - GLM-4-9B is the open-source version of the latest generation of pre-trained models in the GLM-4 series launched by Zhipu AI. In the evaluation of data sets in semantics, mathematics, reasoning, code, and knowledge, GLM-4-9B and its human preference-aligned version GLM-4-9B-Chat have shown superior performance beyond Llama-3-8B. In addition to multi-round conversations, GLM-4-9B-Chat also has advanced features such as web browsing, code execution, custom tool calls (Function Call), and long text reasoning (supporting up to 128K context). More details about model can be found in [model card](https://huggingface.co/THUDM/glm-4-9b-chat/blob/main/README_en.md), [technical report](https://arxiv.org/pdf/2406.12793) and [repository](https://github.com/THUDM/GLM-4) \n",
    "</details>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "0a9b1fed-d145-4fff-828c-ac93ecce5361",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c67b5403bbd5467e8adf09695683dcb7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Box(children=(Box(children=(Label(value='Language:'), Dropdown(options=('English', 'Chinese', 'Japanese'), val…"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from llm_config import get_llm_selection_widget\n",
    "\n",
    "form, lang, model_id_widget, compression_variant, use_preconverted = get_llm_selection_widget(device=device.value)\n",
    "\n",
    "form"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "906022ec-96bf-41a9-9447-789d2e875250",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selected model qwen2.5-0.5b-instruct with INT4 compression\n"
     ]
    }
   ],
   "source": [
    "model_configuration = model_id_widget.value\n",
    "model_id = model_id_widget.label\n",
    "print(f\"Selected model {model_id} with {compression_variant.value} compression\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "62af3e8a-915a-49b4-8007-803777ba9eaf",
   "metadata": {},
   "source": [
    "## Convert model using Optimum-CLI tool\n",
    "[back to top ⬆️](#Table-of-contents:)\n",
    "\n",
    "🤗 [Optimum Intel](https://huggingface.co/docs/optimum/intel/index) is the interface between the 🤗 [Transformers](https://huggingface.co/docs/transformers/index) and [Diffusers](https://huggingface.co/docs/diffusers/index) libraries and OpenVINO to accelerate end-to-end pipelines on Intel architectures. It provides ease-to-use cli interface for exporting models to [OpenVINO Intermediate Representation (IR)](https://docs.openvino.ai/2024/documentation/openvino-ir-format.html) format.\n",
    "\n",
    "<details>\n",
    "  <summary><b>Click here to read more about Optimum CLI usage</b></summary>\n",
    "\n",
    "The command bellow demonstrates basic command for model export with `optimum-cli`\n",
    "\n",
    "```\n",
    "optimum-cli export openvino --model <model_id_or_path> --task <task> <out_dir>\n",
    "```\n",
    "\n",
    "where `--model` argument is model id from HuggingFace Hub or local directory with model (saved using `.save_pretrained` method), `--task ` is one of [supported task](https://huggingface.co/docs/optimum/exporters/task_manager) that exported model should solve. For LLMs it is recommended to use `text-generation-with-past`. If model initialization requires to use remote code, `--trust-remote-code` flag additionally should be passed.\n",
    "</details>\n",
    "\n",
    "### Weights Compression using Optimum-CLI\n",
    "[back to top ⬆️](#Table-of-contents:)\n",
    "\n",
    "You can also apply fp16, 8-bit or 4-bit weight compression on the Linear, Convolutional and Embedding layers when exporting your model with the CLI. \n",
    "<details>\n",
    "  <summary><b>Click here to read more about weights compression with Optimum CLI</b></summary>\n",
    "\n",
    "Setting `--weight-format` to respectively fp16, int8 or int4. This type of optimization allows to reduce the memory footprint and inference latency.\n",
    "By default the quantization scheme for int8/int4 will be [asymmetric](https://github.com/openvinotoolkit/nncf/blob/develop/docs/compression_algorithms/Quantization.md#asymmetric-quantization), to make it [symmetric](https://github.com/openvinotoolkit/nncf/blob/develop/docs/compression_algorithms/Quantization.md#symmetric-quantization) you can add `--sym`.\n",
    "\n",
    "For INT4 quantization you can also specify the following arguments :\n",
    "- The `--group-size` parameter will define the group size to use for quantization, -1 it will results in per-column quantization.\n",
    "- The `--ratio` parameter controls the ratio between 4-bit and 8-bit quantization. If set to 0.9, it means that 90% of the layers will be quantized to int4 while 10% will be quantized to int8.\n",
    "\n",
    "Smaller group_size and ratio values usually improve accuracy at the sacrifice of the model size and inference latency.\n",
    "You can enable AWQ to be additionally applied during model export with INT4 precision using `--awq` flag and providing dataset name with `--dataset`parameter (e.g. `--dataset wikitext2`)\n",
    "\n",
    ">**Note**: Applying AWQ requires significant memory and time.\n",
    "\n",
    ">**Note**: It is possible that there will be no matching patterns in the model to apply AWQ, in such case it will be skipped.\n",
    "</details>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "c4ef9112",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "✅ INT4 qwen2.5-0.5b-instruct model already converted and can be found in qwen2.5/INT4_compressed_weights\n"
     ]
    }
   ],
   "source": [
    "from llm_config import convert_and_compress_model\n",
    "\n",
    "model_dir = convert_and_compress_model(model_id, model_configuration, compression_variant.value, use_preconverted.value)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "671a17d4",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "Let's compare model size for different compression types"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "281f1d07-998e-4e13-ba95-0264564ede82",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Size of model with INT4 compressed weights is 306.44 MB\n",
      "Size of model with INT4-NPU compressed weights is 301.69 MB\n"
     ]
    }
   ],
   "source": [
    "from llm_config import compare_model_size\n",
    "\n",
    "compare_model_size(model_dir)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "f7f63327-f0f5-4e2d-bfc2-0f764f8c19a8",
   "metadata": {},
   "source": [
    "## Instantiate pipeline with OpenVINO Generate API\n",
    "[back to top ⬆️](#Table-of-contents:)\n",
    "\n",
    "[OpenVINO Generate API](https://github.com/openvinotoolkit/openvino.genai/blob/master/src/README.md) can be used to create pipelines to run an inference with OpenVINO Runtime. \n",
    "\n",
    "Firstly we need to create a pipeline with `LLMPipeline`. `LLMPipeline` is the main object used for text generation using LLM in OpenVINO GenAI API. You can construct it straight away from the folder with the converted model. We will provide directory with model and device for `LLMPipeline`. Then we run `generate` method and get the output in text format.\n",
    "Additionally, we can configure parameters for decoding. We can create the default config with `ov_genai.GenerationConfig()`, setup parameters, and apply the updated version with `set_generation_config(config)` or put config directly to `generate()`. It's also possible to specify the needed options just as inputs in the `generate()` method, as shown below, e.g. we can add `max_new_tokens` to stop generation if a specified number of tokens is generated and the end of generation is not reached. We will discuss some of the available generation parameters more deeply later.  Generation process for long response may be time consuming, for accessing partial result as soon as it is generated without waiting when whole process finished, Streaming API can be used. Token streaming is the mode in which the generative system returns the tokens one by one as the model generates them. This enables showing progressive generations to the user rather than waiting for the whole generation. Streaming is an essential aspect of the end-user experience as it reduces latency, one of the most critical aspects of a smooth experience. In code below, we implement simple streamer for printing output result. For more advanced streamer example please check openvino.genai [sample](https://github.com/openvinotoolkit/openvino.genai/tree/master/samples/python/multinomial_causal_lm)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "7a041101-7336-40fd-96c9-cd298015a0f3",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading model from qwen2.5/INT4_compressed_weights\n",
      "\n",
      "Input text: The Sun is yellow bacause\n",
      " of its own light. The Sun is blue because of its own light. The Sun is red because of its own light. The Sun is green because of its own light. The Sun is blue because of its own light. The Sun is red because of its own light. The Sun is green because of its own light. The Sun is yellow because of its own light. The Sun is blue because of its own light. The Sun is red because of its own light. The Sun is green because"
     ]
    },
    {
     "data": {
      "text/plain": [
       "' of its own light. The Sun is blue because of its own light. The Sun is red because of its own light. The Sun is green because of its own light. The Sun is blue because of its own light. The Sun is red because of its own light. The Sun is green because of its own light. The Sun is yellow because of its own light. The Sun is blue because of its own light. The Sun is red because of its own light. The Sun is green because'"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import openvino_genai as ov_genai\n",
    "\n",
    "print(f\"Loading model from {model_dir}\\n\")\n",
    "\n",
    "\n",
    "pipe = ov_genai.LLMPipeline(str(model_dir), device.value)\n",
    "\n",
    "generation_config = ov_genai.GenerationConfig()\n",
    "generation_config.max_new_tokens = 100\n",
    "\n",
    "\n",
    "def streamer(subword):\n",
    "    print(subword, end=\"\", flush=True)\n",
    "    # Return flag corresponds whether generation should be stopped.\n",
    "    # False means continue generation.\n",
    "    return False\n",
    "\n",
    "\n",
    "input_prompt = \"The Sun is yellow bacause\"\n",
    "print(f\"Input text: {input_prompt}\")\n",
    "pipe.generate(input_prompt, generation_config, streamer)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "24d622d0-be46-47c0-a762-88cb50ab15a9",
   "metadata": {},
   "source": [
    "## Run Chatbot\n",
    "[back to top ⬆️](#Table-of-contents:)\n",
    "\n",
    "Now, when model created, we can setup Chatbot interface using [Gradio](https://www.gradio.app/).\n",
    "\n",
    "<details>\n",
    "  <summary><b>Click here to see how pipeline works</b></summary>\n",
    "\n",
    "The diagram below illustrates how the chatbot pipeline works\n",
    "\n",
    "![llm_diagram](https://github.com/user-attachments/assets/9c9b56e1-01a6-48d8-aa46-222a88e25066)\n",
    "\n",
    "As you can see, user input question passed via tokenizer to apply chat-specific formatting (chat template) and turn the provided string into the numeric format. [OpenVINO Tokenizers](https://github.com/openvinotoolkit/openvino_tokenizers) are used for these purposes inside `LLMPipeline`. You can find more detailed info about  tokenization theory and OpenVINO Tokenizers in this [tutorial](https://github.com/openvinotoolkit/openvino_notebooks/blob/latest/notebooks/openvino-tokenizers/openvino-tokenizers.ipynb). Then tokenized input passed to LLM for making prediction of next token probability. The way the next token will be selected over predicted probabilities is driven by the selected decoding methodology.  You can find more information about the most popular decoding methods in this [blog](https://huggingface.co/blog/how-to-generate). The sampler's goal is to select the next token id is driven by generation configuration. Next, we apply stop generation condition to check the generation is finished or not (e.g. if we reached the maximum new generated tokens or the next token id equals to end of the generation).  If the end of the generation is not reached, then new generated token id is used as the next iteration input, and the generation cycle repeats until the condition is not met. When stop generation criteria are met, then OpenVINO Detokenizer decodes generated token ids to text answer. \n",
    "\n",
    "The difference between chatbot and instruction-following pipelines is that the model should have \"memory\" to find correct answers on the chain of connected questions. OpenVINO GenAI uses `KVCache` representation for maintain a history of conversation. By default, `LLMPipeline` resets `KVCache` after each `generate` call. To keep conversational history, we should move LLMPipeline to chat mode using `start_chat()` method.\n",
    "\n",
    "More info about OpenVINO LLM inference can be found in [LLM Inference Guide](https://docs.openvino.ai/2024/learn-openvino/llm_inference_guide.html)\n",
    "</details>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "725544ea-05ec-40d7-bbbc-1dc87cf57d04",
   "metadata": {},
   "source": [
    "### Advanced generation options\n",
    "[back to top ⬆️](#Table-of-contents:)\n",
    "\n",
    "> **Note**: NPU inference pipeline does not support providing advanced options at this moment, they will be disabled if NPU selected\n",
    "\n",
    "<details>\n",
    "  <summary><b>Click here to see detailed description of advanced options</b></summary>\n",
    "\n",
    "There are several parameters that can control text generation quality, \n",
    "  * `Temperature` is a parameter used to control the level of creativity in AI-generated text. By adjusting the `temperature`, you can influence the AI model's probability distribution, making the text more focused or diverse.  \n",
    "  Consider the following example: The AI model has to complete the sentence \"The cat is ____.\" with the following token probabilities:  \n",
    "\n",
    "    playing: 0.5  \n",
    "    sleeping: 0.25  \n",
    "    eating: 0.15  \n",
    "    driving: 0.05  \n",
    "    flying: 0.05  \n",
    "\n",
    "    - **Low temperature** (e.g., 0.2): The AI model becomes more focused and deterministic, choosing tokens with the highest probability, such as \"playing.\"  \n",
    "    - **Medium temperature** (e.g., 1.0): The AI model maintains a balance between creativity and focus, selecting tokens based on their probabilities without significant bias, such as \"playing,\" \"sleeping,\" or \"eating.\"  \n",
    "    - **High temperature** (e.g., 2.0): The AI model becomes more adventurous, increasing the chances of selecting less likely tokens, such as \"driving\" and \"flying.\"\n",
    "  * `Top-p`, also known as nucleus sampling, is a parameter used to control the range of tokens considered by the AI model based on their cumulative probability. By adjusting the `top-p` value, you can influence the AI model's token selection, making it more focused or diverse.\n",
    "  Using the same example with the cat, consider the following top_p settings:  \n",
    "    - **Low top_p** (e.g., 0.5): The AI model considers only tokens with the highest cumulative probability, such as \"playing.\"  \n",
    "    - **Medium top_p** (e.g., 0.8): The AI model considers tokens with a higher cumulative probability, such as \"playing,\" \"sleeping,\" and \"eating.\"  \n",
    "    - **High top_p** (e.g., 1.0): The AI model considers all tokens, including those with lower probabilities, such as \"driving\" and \"flying.\" \n",
    "  * `Top-k` is an another popular sampling strategy. In comparison with Top-P, which chooses from the smallest possible set of words whose cumulative probability exceeds the probability P, in Top-K sampling K most likely next words are filtered and the probability mass is redistributed among only those K next words. In our example with cat, if k=3, then only \"playing\", \"sleeping\" and \"eating\" will be taken into account as possible next word.\n",
    "  * `Repetition Penalty` This parameter can help penalize tokens based on how frequently they occur in the text, including the input prompt. A token that has already appeared five times is penalized more heavily than a token that has appeared only one time. A value of 1 means that there is no penalty and values larger than 1 discourage repeated tokens.\n",
    "</details>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "01f8f7f8-072e-45dc-b7c9-18d8c3c47754",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "if not Path(\"gradio_helper_genai.py\").exists():\n",
    "    r = requests.get(url=\"https://raw.githubusercontent.com/openvinotoolkit/openvino_notebooks/latest/notebooks/llm-chatbot/gradio_helper_genai.py\")\n",
    "    open(\"gradio_helper_genai.py\", \"w\").write(r.text)\n",
    "\n",
    "from gradio_helper_genai import make_demo\n",
    "\n",
    "demo = make_demo(pipe, model_configuration, model_id, lang.value, device.value == \"NPU\")\n",
    "\n",
    "try:\n",
    "    demo.launch(debug=True)\n",
    "except Exception:\n",
    "    demo.launch(debug=True, share=True)\n",
    "# If you are launching remotely, specify server_name and server_port\n",
    "# EXAMPLE: `demo.launch(server_name='your server name', server_port='server port in int')`\n",
    "# To learn more please refer to the Gradio docs: https://gradio.app/docs/"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.4"
  },
  "openvino_notebooks": {
   "imageUrl": "https://user-images.githubusercontent.com/29454499/255799218-611e7189-8979-4ef5-8a80-5a75e0136b50.png",
   "tags": {
    "categories": [
     "Model Demos",
     "AI Trends"
    ],
    "libraries": [],
    "other": [
     "LLM"
    ],
    "tasks": [
     "Text Generation",
     "Conversational"
    ]
   }
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {
     "224609482cef42d18856afc169f73971": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "2db4b2e30a784589a4b699653454cf93": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "DropdownModel",
      "state": {
       "_options_labels": [
        "qwen2.5-0.5b-instruct",
        "tiny-llama-1b-chat",
        "llama-3.2-1b-instruct",
        "llama-3.2-3b-instruct",
        "qwen2.5-1.5b-instruct",
        "gemma-2b-it",
        "gemma-2-2b-it",
        "red-pajama-3b-chat",
        "qwen2.5-3b-instruct",
        "minicpm3-4b",
        "qwen2.5-7b-instruct",
        "gemma-7b-it",
        "gemma-2-9b-it",
        "llama-2-chat-7b",
        "llama-3-8b-instruct",
        "llama-3.1-8b-instruct",
        "mistral-7b-instruct",
        "zephyr-7b-beta",
        "notus-7b-v1",
        "neural-chat-7b-v3-3",
        "phi-3-mini-instruct",
        "phi-3.5-mini-instruct",
        "qwen2.5-14b-instruct"
       ],
       "index": 0,
       "layout": "IPY_MODEL_743be0e91d994804af1527013202ef9c",
       "style": "IPY_MODEL_874757d4dca7445eb7e0ebbf05760ef6"
      }
     },
     "35145f4c96224bca92c86324a462bdc9": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "LabelModel",
      "state": {
       "layout": "IPY_MODEL_f38a49dc2ce4410583e2cff6d245c516",
       "style": "IPY_MODEL_7e7e5661615d4adf958567edd8f749bd",
       "value": "Language:"
      }
     },
     "36a23d4e0bfd4e9086725fff37a2351a": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "3acab846363844bbbb3a9cc37cfe0f59": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "LabelModel",
      "state": {
       "layout": "IPY_MODEL_e479517b942043718dfad27670bcd359",
       "style": "IPY_MODEL_dfdc609844be4ac093304d9555c25cc5",
       "value": "Model:"
      }
     },
     "3fcb108534f344ca822b7ef60c764652": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "LabelModel",
      "state": {
       "layout": "IPY_MODEL_7395325b5b6d4374856a6f771dd7e43b",
       "style": "IPY_MODEL_c2ceb3b6cc224f37b9030e2a7f965270",
       "value": "Compression:"
      }
     },
     "49ee3dbd07c94d26a42925f80e7a160f": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "650181049bdb4ffcb84afc9aee554261": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "DropdownModel",
      "state": {
       "_options_labels": [
        "CPU",
        "AUTO"
       ],
       "description": "Device:",
       "index": 0,
       "layout": "IPY_MODEL_49ee3dbd07c94d26a42925f80e7a160f",
       "style": "IPY_MODEL_7ff49c71220246329555706db225447d"
      }
     },
     "68c303387c0b4f8294cefe870fb5b941": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "CheckboxStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "6c076a58bb044a1988fbeb4c24bbc822": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "DropdownModel",
      "state": {
       "_options_labels": [
        "English",
        "Chinese",
        "Japanese"
       ],
       "index": 0,
       "layout": "IPY_MODEL_36a23d4e0bfd4e9086725fff37a2351a",
       "style": "IPY_MODEL_ab83c42251ac4fd2a6b1492a472ca34e"
      }
     },
     "71703e3b27c34711892df1d661c0de9f": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "7395325b5b6d4374856a6f771dd7e43b": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "743be0e91d994804af1527013202ef9c": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "74b02264db1348be84f10324930772cb": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "75971ce865724d61b0b8c9815a2d71fe": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "7e7e5661615d4adf958567edd8f749bd": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "LabelStyleModel",
      "state": {
       "description_width": "",
       "font_family": null,
       "font_size": null,
       "font_style": null,
       "font_variant": null,
       "font_weight": null,
       "text_color": null,
       "text_decoration": null
      }
     },
     "7ff49c71220246329555706db225447d": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "81e48003c0ab4695a2125b7f439ddb0e": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "8484b3c285d24e769886a8d623751e61": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "DropdownModel",
      "state": {
       "_options_labels": [
        "INT4",
        "INT4-AWQ",
        "INT4-NPU",
        "INT8",
        "FP16"
       ],
       "index": 0,
       "layout": "IPY_MODEL_a30cab2d17b84948b32fd209aea44d87",
       "style": "IPY_MODEL_71703e3b27c34711892df1d661c0de9f"
      }
     },
     "874757d4dca7445eb7e0ebbf05760ef6": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "8a192162dc6346848d9642285fa512ca": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "BoxModel",
      "state": {
       "children": [
        "IPY_MODEL_985afd9d4f6d4dd99b0f0d0e208fc1f4",
        "IPY_MODEL_a3e9f4a29a674ed5be3f837d10c1dc42"
       ],
       "layout": "IPY_MODEL_fd3727bd7d4649e1aa6fb252e1e4db65"
      }
     },
     "985afd9d4f6d4dd99b0f0d0e208fc1f4": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "LabelModel",
      "state": {
       "layout": "IPY_MODEL_75971ce865724d61b0b8c9815a2d71fe",
       "style": "IPY_MODEL_e67b277f19c249eab61593e5bd65424a",
       "value": "Use preconverted models:"
      }
     },
     "a076c4d675f74b67babd15d94cdbbd61": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "border_bottom": "solid 1px",
       "border_left": "solid 1px",
       "border_right": "solid 1px",
       "border_top": "solid 1px",
       "display": "flex",
       "flex_flow": "column",
       "padding": "1%",
       "width": "30%"
      }
     },
     "a30cab2d17b84948b32fd209aea44d87": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "a3e9f4a29a674ed5be3f837d10c1dc42": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "CheckboxModel",
      "state": {
       "disabled": false,
       "layout": "IPY_MODEL_224609482cef42d18856afc169f73971",
       "style": "IPY_MODEL_68c303387c0b4f8294cefe870fb5b941",
       "value": true
      }
     },
     "a86a9e42953e4973ae5ede347f9f9b0b": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "ab83c42251ac4fd2a6b1492a472ca34e": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "c2ceb3b6cc224f37b9030e2a7f965270": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "LabelStyleModel",
      "state": {
       "description_width": "",
       "font_family": null,
       "font_size": null,
       "font_style": null,
       "font_variant": null,
       "font_weight": null,
       "text_color": null,
       "text_decoration": null
      }
     },
     "c2d5d09c906c4905a7f4041310c836b9": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "BoxModel",
      "state": {
       "children": [
        "IPY_MODEL_3acab846363844bbbb3a9cc37cfe0f59",
        "IPY_MODEL_2db4b2e30a784589a4b699653454cf93"
       ],
       "layout": "IPY_MODEL_74b02264db1348be84f10324930772cb"
      }
     },
     "c67b5403bbd5467e8adf09695683dcb7": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "BoxModel",
      "state": {
       "children": [
        "IPY_MODEL_dd288be95cdd4d89afba1adbebf67d08",
        "IPY_MODEL_c2d5d09c906c4905a7f4041310c836b9",
        "IPY_MODEL_d7a95d1568854888b06d834124578cd8",
        "IPY_MODEL_8a192162dc6346848d9642285fa512ca"
       ],
       "layout": "IPY_MODEL_a076c4d675f74b67babd15d94cdbbd61"
      }
     },
     "d7a95d1568854888b06d834124578cd8": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "BoxModel",
      "state": {
       "children": [
        "IPY_MODEL_3fcb108534f344ca822b7ef60c764652",
        "IPY_MODEL_8484b3c285d24e769886a8d623751e61"
       ],
       "layout": "IPY_MODEL_81e48003c0ab4695a2125b7f439ddb0e"
      }
     },
     "dd288be95cdd4d89afba1adbebf67d08": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "BoxModel",
      "state": {
       "children": [
        "IPY_MODEL_35145f4c96224bca92c86324a462bdc9",
        "IPY_MODEL_6c076a58bb044a1988fbeb4c24bbc822"
       ],
       "layout": "IPY_MODEL_a86a9e42953e4973ae5ede347f9f9b0b"
      }
     },
     "dfdc609844be4ac093304d9555c25cc5": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "LabelStyleModel",
      "state": {
       "description_width": "",
       "font_family": null,
       "font_size": null,
       "font_style": null,
       "font_variant": null,
       "font_weight": null,
       "text_color": null,
       "text_decoration": null
      }
     },
     "e479517b942043718dfad27670bcd359": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "e67b277f19c249eab61593e5bd65424a": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "LabelStyleModel",
      "state": {
       "description_width": "",
       "font_family": null,
       "font_size": null,
       "font_style": null,
       "font_variant": null,
       "font_weight": null,
       "text_color": null,
       "text_decoration": null
      }
     },
     "f38a49dc2ce4410583e2cff6d245c516": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "fd3727bd7d4649e1aa6fb252e1e4db65": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {}
     }
    },
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}