{"cells":[{"cell_type":"markdown","metadata":{"id":"7PJjHfmPfuVA"},"source":["# Partager ses démos avec d'autres"]},{"cell_type":"markdown","metadata":{"id":"jRWYhCTHfuVC"},"source":["Installez les bibliothèques 🤗 Transformers et 🤗 Gradio pour exécuter ce *notebook*."]},{"cell_type":"code","execution_count":null,"metadata":{"id":"nhJZ6vw1fuVD"},"outputs":[],"source":["!pip install datasets transformers[sentencepiece]\n","!pip install gradio"]},{"cell_type":"code","source":[],"metadata":{"id":"nrI9JVsynK2f"},"execution_count":null,"outputs":[]},{"cell_type":"code","execution_count":null,"metadata":{"id":"9I2ia0R1fuVE"},"outputs":[],"source":["import gradio as gr\n","\n","title = \"Poser une question (en anglais) à Rick\"\n","description = \"\"\"\n","Le bot a été entraîné à répondre à des questions basées sur les dialogues de Rick et Morty (en anglais). Demandez à Rick ce que vous voulez !\n","<img src=\"https://huggingface.co/spaces/course-demos/Rick_and_Morty_QA/resolve/main/rick.png\" width=200px>\n","\"\"\"\n","\n","article = \"Consultez [le bot original Rick et Morty](https://huggingface.co/spaces/kingabzpro/Rick_and_Morty_Bot) sur lequel cette démo est basée.\"\n","\n","from transformers import AutoModelForCausalLM, AutoTokenizer\n","import torch\n","\n","tokenizer = AutoTokenizer.from_pretrained(\"ericzhou/DialoGPT-Medium-Rick_v2\")\n","model = AutoModelForCausalLM.from_pretrained(\"ericzhou/DialoGPT-Medium-Rick_v2\")\n","\n","def predict(input, history=[]):\n","    # tokenizer la nouvelle phrase d'entrée\n","    new_user_input_ids = tokenizer.encode(input + tokenizer.eos_token, return_tensors='pt')\n","\n","    # ajouter les nouveaux tokens d'entrée de l'utilisateur à l'historique de chat\n","    bot_input_ids = torch.cat([torch.LongTensor(history), new_user_input_ids], dim=-1)\n","\n","    # générer une réponse \n","    history = model.generate(bot_input_ids, max_length=1000, pad_token_id=tokenizer.eos_token_id).tolist()\n","\n","    # convertit les tokens en texte, puis divise les réponses dans le bon format.\n","    response = tokenizer.decode(history[0]).split(\"<|endoftext|>\")\n","    response = [(response[i], response[i+1]) for i in range(0, len(response)-1, 2)]  # convertir en tuples de liste\n","    return response, history\n","\n","gr.Interface(\n","    fn=predict,\n","    inputs=\"textbox\",\n","    outputs=\"text\",\n","    title=title,\n","    description=description,\n","    article=article,\n","    examples=[[\"What are you doing?\"], [\"Where should we time travel to?\"]],\n",").launch()"]},{"cell_type":"code","source":[],"metadata":{"id":"XDfMUCtb6Stg"},"execution_count":null,"outputs":[]},{"cell_type":"code","execution_count":null,"metadata":{"id":"-FIvt7difuVG"},"outputs":[],"source":["# Vous devez récupérer le fichier pytorch_model.bin ici https://huggingface.co/spaces/course-demos/Sketch-Recognition/blob/main/pytorch_model.bin\n","\n","import torch\n","import gradio as gr\n","from torch import nn\n","import requests\n","from google.colab import drive\n","drive.mount('/content/MyDrive/pytorch_model.bin')\n","\n","LABELS = requests.get(\"https://huggingface.co/spaces/course-demos/Sketch-Recognition/raw/main/class_names.txt\").text.replace(\"\\n\",\"\").split(\"\\r\")\n","\n","model = nn.Sequential(\n","    nn.Conv2d(1, 32, 3, padding=\"same\"),\n","    nn.ReLU(),\n","    nn.MaxPool2d(2),\n","    nn.Conv2d(32, 64, 3, padding=\"same\"),\n","    nn.ReLU(),\n","    nn.MaxPool2d(2),\n","    nn.Conv2d(64, 128, 3, padding=\"same\"),\n","    nn.ReLU(),\n","    nn.MaxPool2d(2),\n","    nn.Flatten(),\n","    nn.Linear(1152, 256),\n","    nn.ReLU(),\n","    nn.Linear(256, len(LABELS)),\n",")\n","state_dict = torch.load(\"pytorch_model.bin\", map_location=\"cpu\")\n","model.load_state_dict(state_dict, strict=False)\n","model.eval()\n","\n","\n","def predict(im):\n","    x = torch.tensor(im, dtype=torch.float32).unsqueeze(0).unsqueeze(0) / 255.0\n","    with torch.no_grad():\n","        out = model(x)\n","    probabilities = torch.nn.functional.softmax(out[0], dim=0)\n","    values, indices = torch.topk(probabilities, 5)\n","    return {LABELS[i]: v.item() for i, v in zip(indices, values)}\n","\n","\n","interface = gr.Interface(\n","    predict,\n","    inputs=\"sketchpad\",\n","    outputs=\"label\",\n","    theme=\"huggingface\",\n","    title=\"Reconnaissance de croquis\",\n","    description=\"Qui veut jouer au Pictionary ? Dessinez un objet courant comme une pelle ou un ordinateur portable, et l'algorithme le devinera en temps réel !\",\n","    article=\"<p style='text-align: center'>Reconnaissance de croquis | Modèle de démonstration</p>\",\n","    live=True,\n",")\n","interface.launch(share=True)"]}],"metadata":{"colab":{"provenance":[],"collapsed_sections":[]},"kernelspec":{"display_name":"Python 3","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":"3.8.5"}},"nbformat":4,"nbformat_minor":0}