main.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load the modules, classes and configs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import torch \n",
    "import torch.optim as optim\n",
    "from torch.utils.data import DataLoader\n",
    "from torchvision.transforms import Compose, ToTensor, Normalize, Resize\n",
    "\n",
    "from sde import VPSDE\n",
    "from models.unet import UNet\n",
    "from loss import get_sde_loss_fn\n",
    "from configs.config import CFGS\n",
    "from train import Step_by_Step\n",
    "from dataset import myDataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Setup the stochastic differential equation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the sde\n",
    "vpsde = VPSDE(N=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Setup the score model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the models\n",
    "model_config = CFGS[\"model\"]\n",
    "model_xy = UNet(**model_config)\n",
    "model_yx = UNet(**model_config)\n",
    "\n",
    "models = (model_xy, model_yx)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Setup the loss function for the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the loss function\n",
    "loss_fn = get_sde_loss_fn(sde=vpsde)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Setup the optimizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the optimizer\n",
    "optimizer_xy = optim.Adam(model_xy.parameters(), lr=CFGS[\"optim\"][\"lr\"], betas=(CFGS[\"optim\"][\"beta1\"], 0.999), \n",
    "                       eps=CFGS[\"optim\"][\"eps\"], weight_decay=CFGS[\"optim\"][\"weight_decay\"])\n",
    "\n",
    "optimizer_yx = optim.Adam(model_yx.parameters(), lr=CFGS[\"optim\"][\"lr\"], betas=(CFGS[\"optim\"][\"beta1\"], 0.999), \n",
    "                       eps=CFGS[\"optim\"][\"eps\"], weight_decay=CFGS[\"optim\"][\"weight_decay\"])\n",
    "\n",
    "optimizers = (optimizer_xy, optimizer_yx)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Setup the training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Instantiate the step by step class\n",
    "sbs = Step_by_Step(sde=vpsde, models=models, loss_fn=loss_fn, optimizers=optimizers, config=CFGS)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load the dataset into dataloaders"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_dir = \"./datasets/horse2zebra\"\n",
    "img_size = CFGS[\"model\"][\"image_size\"]\n",
    "\n",
    "transforms = Compose(\n",
    "    [Resize((img_size, img_size)),\n",
    "     ToTensor(),\n",
    "     Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),\n",
    "    ]\n",
    ")\n",
    "\n",
    "train_dataset = myDataset(\n",
    "    root_A= os.path.join(dataset_dir, \"trainA\"),\n",
    "    root_B= os.path.join(dataset_dir, \"trainB\"),\n",
    "    transform=transforms\n",
    ")\n",
    "\n",
    "# validation_dataset = myDataset(\n",
    "#     root_A= os.path.join(dataset_dir, \"valA\"),\n",
    "#     root_B= os.path.join(dataset_dir, \"valB\"),\n",
    "#     transform=transforms\n",
    "\n",
    "# )\n",
    "\n",
    "train_loader = DataLoader(\n",
    "    train_dataset,\n",
    "    batch_size=CFGS[\"training\"][\"batch_size\"],\n",
    "    num_workers = 4,\n",
    "    pin_memory=True,\n",
    "    drop_last=True,\n",
    "    shuffle=True\n",
    ")\n",
    "\n",
    "# val_loader = DataLoader(\n",
    "#     validation_dataset,\n",
    "#     batch_size=CFGS[\"training\"][\"batch_size\"],\n",
    "#     num_workers = 4,\n",
    "#     pin_memory=True\n",
    "# )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set the data loaders\n",
    "sbs.set_loaders(train_loader=train_loader)\n",
    "#sbs.set_tensorboard(\"train\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the model from the previous checkpoint\n",
    "#sbs.load_checkpoint(\"./checkpoints/checkpoint.pth\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Train Epoch: 402: 100%|██████████| 128/128 [05:24<00:00,  2.53s/batch]\n",
      "Train Epoch: 403: 100%|██████████| 128/128 [05:23<00:00,  2.53s/batch]\n",
      "Train Epoch: 404: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 405: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 406: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 407: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 408: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 409: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 410: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 411: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 412: 100%|██████████| 128/128 [05:23<00:00,  2.52s/batch]\n",
      "Train Epoch: 413: 100%|██████████| 128/128 [05:23<00:00,  2.52s/batch]\n",
      "Train Epoch: 414: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 415: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 416: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 417: 100%|██████████| 128/128 [05:23<00:00,  2.52s/batch]\n",
      "Train Epoch: 418: 100%|██████████| 128/128 [05:23<00:00,  2.52s/batch]\n",
      "Train Epoch: 419: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 420: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 421: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epochs Completed: 421\n",
      "Current Time: 23:56:49\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Train Epoch: 422: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 423: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 424: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 425: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 426: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 427: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 428: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 429: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 430: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 431: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 432: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 433: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 434: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 435: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 436: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 437: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 438: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 439: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 440: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 441: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epochs Completed: 441\n",
      "Current Time: 01:43:53\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Train Epoch: 442: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 443: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 444: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 445: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 446: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 447: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 448: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 449: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 450: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 451: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 452: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 453: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 454: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 455: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 456: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 457: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 458: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 459: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 460: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 461: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epochs Completed: 461\n",
      "Current Time: 03:30:57\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Train Epoch: 462: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 463: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 464: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 465: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 466: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 467: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 468: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 469: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 470: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 471: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 472: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 473: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 474: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 475: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 476: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 477: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 478: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 479: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 480: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 481: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epochs Completed: 481\n",
      "Current Time: 05:18:15\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Train Epoch: 482: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 483: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 484: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 485: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 486: 100%|██████████| 128/128 [05:21<00:00,  2.51s/batch]\n",
      "Train Epoch: 487: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 488: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 489: 100%|██████████| 128/128 [05:20<00:00,  2.51s/batch]\n",
      "Train Epoch: 490: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 491: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 492: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 493: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 494: 100%|██████████| 128/128 [05:22<00:00,  2.52s/batch]\n",
      "Train Epoch: 495: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 496: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 497: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 498: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 499: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 500: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n",
      "Train Epoch: 501: 100%|██████████| 128/128 [05:20<00:00,  2.50s/batch]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epochs Completed: 501\n",
      "Current Time: 07:05:25\n"
     ]
    }
   ],
   "source": [
    "sbs.train(501)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sbs.plot_losses()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.07463913026731461"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sbs.losses[-1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model inference"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_dir = \"./datasets/inference\"\n",
    "img_size = CFGS[\"model\"][\"image_size\"]\n",
    "\n",
    "transforms = Compose(\n",
    "    [Resize((img_size, img_size)),\n",
    "     ToTensor(),\n",
    "     Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),\n",
    "    ]\n",
    ")\n",
    "\n",
    "infer_dataset = myDataset(\n",
    "    root_A= os.path.join(dataset_dir, \"horses\"),\n",
    "    root_B= os.path.join(dataset_dir, \"zebra\"),\n",
    "    transform=transforms\n",
    ")\n",
    "\n",
    "inference_loader = DataLoader(infer_dataset, shuffle=True, batch_size=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "condition = next(iter(inference_loader))[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 288x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sbs.plot_samples(condition)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "samples_x = sbs.translate(target_domain=\"x\", condition=condition, num_steps=1500)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 288x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sbs.plot_samples(samples_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "mykernel",
   "language": "python",
   "name": "mykernel"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}