slime-lightning-old.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "HiGHZ1qiq_9q"
      },
      "outputs": [],
      "source": [
        "%load_ext autoreload\n",
        "%autoreload 2"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "OIu3N4tGnEgv"
      },
      "outputs": [],
      "source": [
        "should_skip = True"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "cellView": "form",
        "id": "n0vWJ_IynEgv"
      },
      "outputs": [],
      "source": [
        "#@title Install libraries\n",
        "if not should_skip:\n",
        "  !pip install -U -qq git+https://github.com/huggingface/diffusers.git\n",
        "  !pip install -qq accelerate transformers ftfy\n",
        "  !pip install -q gpustat gdown\n",
        "  !pip install -q pytorch_lightning"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "cellView": "form",
        "id": "vey7mPaRnEgw"
      },
      "outputs": [],
      "source": [
        "#@title Import required libraries\n",
        "import argparse\n",
        "import itertools\n",
        "import math\n",
        "import os\n",
        "import random\n",
        "\n",
        "import numpy as np\n",
        "import torch\n",
        "import torch.nn.functional as F\n",
        "import torchvision.transforms.functional as TVF\n",
        "import torch.utils.checkpoint\n",
        "from torch.utils.data import Dataset\n",
        "\n",
        "import PIL\n",
        "from accelerate import Accelerator\n",
        "from accelerate.logging import get_logger\n",
        "from accelerate.utils import set_seed\n",
        "import sys\n",
        "\n",
        "# https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/unet_2d_condition.py#L175-L182\n",
        "from diffusers import AutoencoderKL, DDPMScheduler, PNDMScheduler, StableDiffusionPipeline, UNet2DConditionModel\n",
        "from diffusers.optimization import get_scheduler\n",
        "from diffusers.pipelines.stable_diffusion import StableDiffusionSafetyChecker\n",
        "from PIL import Image\n",
        "from torchvision import transforms\n",
        "from tqdm.auto import tqdm\n",
        "from transformers import CLIPFeatureExtractor, CLIPTextModel, CLIPTokenizer\n",
        "\n",
        "def image_grid(imgs, rows, cols):\n",
        "    assert len(imgs) == rows*cols\n",
        "\n",
        "    w, h = imgs[0].size\n",
        "    grid = Image.new('RGB', size=(cols*w, rows*h))\n",
        "    grid_w, grid_h = grid.size\n",
        "\n",
        "    for i, img in enumerate(imgs):\n",
        "        grid.paste(img, box=(i%cols*w, i//cols*h))\n",
        "    return grid"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "cellView": "form",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TlbAj5_onEgw",
        "outputId": "5d7b3d8b-0f86-456b-e4fa-b1ab2ca50426"
      },
      "outputs": [],
      "source": [
        "if not should_skip:\n",
        "    #@markdown Add here the URLs to the images of the concept you are adding. 3-5 should be fine\n",
        "    train_urls = [\n",
        "        \"https://huggingface.co/datasets/valhalla/images/resolve/main/2.jpeg\",\n",
        "        \"https://huggingface.co/datasets/valhalla/images/resolve/main/3.jpeg\",\n",
        "    ]\n",
        "    test_urls = [\n",
        "        \"https://huggingface.co/datasets/valhalla/images/resolve/main/5.jpeg\",\n",
        "        \"https://huggingface.co/datasets/valhalla/images/resolve/main/6.jpeg\",\n",
        "        ## You can add additional images here\n",
        "    ]\n",
        "#@title Download\n",
        "import requests\n",
        "import glob\n",
        "from io import BytesIO\n",
        "\n",
        "def download_image(url):\n",
        "  try:\n",
        "    response = requests.get(url)\n",
        "  except:\n",
        "    return None\n",
        "  return Image.open(BytesIO(response.content)).convert(\"RGB\")\n",
        "\n",
        "if not should_skip:\n",
        "    for save_path,urls in [(\"./my_concept\",train_urls),(\"./my_concept_test\",test_urls)]:\n",
        "      images = list(filter(None,[download_image(url) for url in urls]))\n",
        "      if not os.path.exists(save_path):\n",
        "        os.mkdir(save_path)\n",
        "      [image.save(f\"{save_path}/{i}.jpeg\") for i, image in enumerate(images)]\n",
        "\n",
        "\n",
        "if not should_skip:\n",
        "    !gdown --id 1I8RMKQhPXntlRKorVNZ6yyNpfaOp8wNv\n",
        "    mask_path=\"./my_concept_gt\"\n",
        "    !mkdir $mask_path/\n",
        "\n",
        "    with open(\"my_concept_gt.zip\",\"r\") as f:\n",
        "        pass\n",
        "    !unzip -o my_concept_gt.zip"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zobTtmaarUxr",
        "outputId": "88ef2419-e38b-460b-f2f8-f6d618d85302"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "[Errno 2] No such file or directory: '/content/'\n",
            "/root/slime-experiments\n"
          ]
        }
      ],
      "source": [
        "%cd /content/\n",
        "sys.path.append(\"/content/slime-experiments/\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "S7EY-HMInhWK",
        "outputId": "bbdf3606-38af-403b-93e2-3e9f5a10ddcf"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Cloning into 'slime-experiments'...\n",
            "remote: Enumerating objects: 218, done.\u001b[K\n",
            "remote: Counting objects: 100% (218/218), done.\u001b[K\n",
            "remote: Compressing objects: 100% (133/133), done.\u001b[K\n",
            "remote: Total 218 (delta 135), reused 165 (delta 82), pack-reused 0\u001b[K\n",
            "Receiving objects: 100% (218/218), 4.64 MiB | 9.63 MiB/s, done.\n",
            "Resolving deltas: 100% (135/135), done.\n",
            "279e51d\n"
          ]
        }
      ],
      "source": [
        "# !rm -rf slime-experiments/\n",
        "# !git clone https://github.com/andrew-healey/slime-experiments\n",
        "# !cd slime-experiments && git rev-parse --short HEAD"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "18OU0iWXnEgw"
      },
      "outputs": [],
      "source": [
        "from slime.slime import SLiME\n",
        "from slime.data.single_class import BinarySegmentationDataset,SegmentationDataModule"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "gjK_JKcsnEgw"
      },
      "outputs": [],
      "source": [
        "train_dataset = BinarySegmentationDataset(\n",
        "    data_root=\"my_concept\",\n",
        "    mask_root=\"my_concept_gt\",\n",
        ")\n",
        "test_dataset = BinarySegmentationDataset(\n",
        "    data_root=\"my_concept_test\",\n",
        "    mask_root=None\n",
        ")\n",
        "\n",
        "data_module = SegmentationDataModule(\n",
        "    train_dataset,\n",
        "    test_dataset,\n",
        "    iters_per_epoch=50,\n",
        "    # batch_size=1,\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5xVozx20nEgw",
        "outputId": "109d1ce9-2c7f-4348-92d7-288ae1714707"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/diffusers/configuration_utils.py:239: FutureWarning: It is deprecated to pass a pretrained model name or path to `from_config`.If you were trying to load a scheduler, please use <class 'diffusers.schedulers.scheduling_ddpm.DDPMScheduler'>.from_pretrained(...) instead. Otherwise, please make sure to pass a configuration dictionary instead. This functionality will be removed in v1.0.0.\n",
            "  deprecate(\"config-passed-as-path\", \"1.0.0\", deprecation_message, standard_warn=False)\n"
          ]
        }
      ],
      "source": [
        "slime = SLiME(\n",
        "    use_self_attn=False,\n",
        ").cuda()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 414,
          "referenced_widgets": [
            "2fe6200bf7064a189e3790f7797c7fbb",
            "d539050cc54f45c288d912144d081d76",
            "5e4c349b29f14b2d85c5e16e78d352ea",
            "817355e86f614485977a900b659a2fed",
            "dc36d39ee4294f8abcc589a88ebc4ceb",
            "6607f73936b44302a81d3f3d3ba28310",
            "39dea76fbc3545658df84b33787383ba",
            "1ef4b2520aae48b3bd830a1d3f179f89",
            "1d50c28ae9b645adb692128f663ee40c",
            "c905811dc695483293ba977f973fef8c",
            "b28003169cce426a875cb82e42cd3c90"
          ]
        },
        "id": "dKcY9CXqnEgw",
        "outputId": "3be2171b-d52d-4163-d7b2-28fd5d29886e"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "GPU available: True (cuda), used: True\n",
            "TPU available: False, using: 0 TPU cores\n",
            "IPU available: False, using: 0 IPUs\n",
            "HPU available: False, using: 0 HPUs\n",
            "`Trainer(limit_train_batches=1.0)` was configured so 100% of the batches per epoch will be used..\n",
            "You are using a CUDA device ('NVIDIA GeForce RTX 3060') that has Tensor Cores. To properly utilize them, you should set `torch.set_float32_matmul_precision('medium' | 'high')` which will trade-off precision for performance. For more details, read https://pytorch.org/docs/stable/generated/torch.set_float32_matmul_precision.html#torch.set_float32_matmul_precision\n",
            "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n",
            "\n",
            "  | Name                   | Type            | Params\n",
            "-----------------------------------------------------------\n",
            "0 | sd                     | StableDiffusion | 900 M \n",
            "1 | cross_layer_multiplier | Multiplier      | 32    \n",
            "2 | self_layer_multiplier  | Multiplier      | 32    \n",
            "3 | cross_map_multiplier   | Multiplier      | 4     \n",
            "4 | pred_map_multiplier    | Multiplier      | 4     \n",
            "5 | cross_norm             | BatchNorm2d     | 10    \n",
            "6 | cross_map_norm         | BatchNorm2d     | 4     \n",
            "  | other params           | n/a             | 2.0 K \n",
            "-----------------------------------------------------------\n",
            "2.1 K     Trainable params\n",
            "900 M     Non-trainable params\n",
            "900 M     Total params\n",
            "3,600.306 Total estimated model params size (MB)\n"
          ]
        },
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "53c135f8dfff4a3f84682d1afb31fc81",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "Training: |          | 0/? [00:00<?, ?it/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "pred -2.94 2.06 ce loss 0.95 mse loss 0.36 dice loss 0.75\n",
            "pred -1.35 0.61 ce loss 0.57 mse loss 0.19 dice loss 0.58\n"
          ]
        }
      ],
      "source": [
        "from pytorch_lightning import Trainer\n",
        "\n",
        "trainer = Trainer(\n",
        "    default_root_dir=\"run\",\n",
        "    limit_train_batches=1.0,\n",
        "    num_sanity_val_steps=0,\n",
        "    max_epochs=50,\n",
        "    enable_checkpointing=False,\n",
        ")\n",
        "\n",
        "trainer.fit(slime, datamodule=data_module)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "UbFajwxfrs0m"
      },
      "outputs": [],
      "source": [
        "slime.cuda()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "YM4p6T8_x6Za",
        "outputId": "e82d4b78-5084-4055-8e37-f5bf8cefaa3d"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x797f6d97aa40>"
            ]
          },
          "execution_count": 47,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.imshow(slime.latest_preds[-1].swapaxes(1,2).reshape((2,2,64,64))[0,1])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 847
        },
        "id": "HQQ3-WNo8B8s",
        "outputId": "4a50b8da-d2fe-45ee-87f1-d5ac8b715749"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "#@title Infer on test set\n",
        "out_mask = None\n",
        "loader = data_module.test_dataloader()\n",
        "for i,batch in enumerate(loader):\n",
        "  out_mask = slime.predict_step(batch,i).cpu().numpy()\n",
        "  break\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "for i in range(len(out_mask)):\n",
        "  plt.imshow(out_mask[i].reshape((64,64)))\n",
        "  plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "_SfOQS29r0q6"
      },
      "outputs": [],
      "source": [
        "!cd /content/slime-experiments/ && git diff"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "gpuType": "T4",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    },
    "orig_nbformat": 4,
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "1d50c28ae9b645adb692128f663ee40c": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "1ef4b2520aae48b3bd830a1d3f179f89": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": "2",
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "2fe6200bf7064a189e3790f7797c7fbb": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_d539050cc54f45c288d912144d081d76",
              "IPY_MODEL_5e4c349b29f14b2d85c5e16e78d352ea",
              "IPY_MODEL_817355e86f614485977a900b659a2fed"
            ],
            "layout": "IPY_MODEL_dc36d39ee4294f8abcc589a88ebc4ceb"
          }
        },
        "39dea76fbc3545658df84b33787383ba": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "5e4c349b29f14b2d85c5e16e78d352ea": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_1ef4b2520aae48b3bd830a1d3f179f89",
            "max": 1,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_1d50c28ae9b645adb692128f663ee40c",
            "value": 0
          }
        },
        "6607f73936b44302a81d3f3d3ba28310": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "817355e86f614485977a900b659a2fed": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_c905811dc695483293ba977f973fef8c",
            "placeholder": "​",
            "style": "IPY_MODEL_b28003169cce426a875cb82e42cd3c90",
            "value": " 0/? [00:00&lt;?, ?it/s, v_num=4]"
          }
        },
        "b28003169cce426a875cb82e42cd3c90": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "c905811dc695483293ba977f973fef8c": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "d539050cc54f45c288d912144d081d76": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_6607f73936b44302a81d3f3d3ba28310",
            "placeholder": "​",
            "style": "IPY_MODEL_39dea76fbc3545658df84b33787383ba",
            "value": "Epoch 499: "
          }
        },
        "dc36d39ee4294f8abcc589a88ebc4ceb": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": "inline-flex",
            "flex": null,
            "flex_flow": "row wrap",
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": "100%"
          }
        }
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}