2024DIS-training.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Config"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from dataclasses import dataclass\n",
    "\n",
    "@dataclass\n",
    "class TrainingConfig:\n",
    "    image_size = 128  # the generated image resolution\n",
    "    train_batch_size = 16\n",
    "    eval_batch_size = 16  # how many images to sample during evaluation\n",
    "    num_epochs = 100\n",
    "    gradient_accumulation_steps = 1\n",
    "    learning_rate = 1e-4\n",
    "    lr_warmup_steps = 500\n",
    "    save_image_epochs = 10\n",
    "    save_model_epochs = 10\n",
    "    mixed_precision = \"fp16\"  # `no` for float32, `fp16` for automatic mixed precision\n",
    "    output_dir = \"flowers_August_100\"  # the model name locally and on the HF Hub\n",
    "\n",
    "    push_to_hub = False  # whether to upload the saved model to the HF Hub\n",
    "    hub_model_id = \"<your-username>/<my-awesome-model>\"  # the name of the repository to create on the HF Hub\n",
    "    hub_private_repo = False\n",
    "    overwrite_output_dir = False  # overwrite the old model when re-running the notebook\n",
    "    seed = 0\n",
    "\n",
    "\n",
    "config = TrainingConfig()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "#!pip install datasets\n",
    "#!pip install torchvision\n",
    "#!pip install diffusers\n",
    "#!pip install torch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/antonia/DIS2024/.venv/lib/python3.10/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "from datasets import load_dataset\n",
    "\n",
    "config.dataset_name = \"nelorth/oxford-flowers\"\n",
    "dataset = load_dataset(config.dataset_name, split=\"train\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Transform the images\n",
    "\n",
    "from torchvision import transforms\n",
    "\n",
    "preprocess = transforms.Compose(\n",
    "    [\n",
    "        transforms.Resize((config.image_size, config.image_size)),\n",
    "        transforms.RandomHorizontalFlip(),\n",
    "        transforms.ToTensor(),\n",
    "        transforms.Normalize([0.5], [0.5]),\n",
    "    ]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def transform(examples):\n",
    "    images = [preprocess(image.convert(\"RGB\")) for image in examples[\"image\"]]\n",
    "    return {\"images\": images}\n",
    "\n",
    "\n",
    "dataset.set_transform(transform)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "\n",
    "train_dataloader = torch.utils.data.DataLoader(dataset, batch_size=config.train_batch_size, shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from diffusers import UNet2DModel\n",
    "\n",
    "model = UNet2DModel(\n",
    "    sample_size=config.image_size,  # the target image resolution\n",
    "    in_channels=3,  # the number of input channels, 3 for RGB images\n",
    "    out_channels=3,  # the number of output channels\n",
    "    layers_per_block=2,  # how many ResNet layers to use per UNet block\n",
    "    block_out_channels=(128, 128, 256, 256, 512, 512),  # the number of output channels for each UNet block\n",
    "    down_block_types=(\n",
    "        \"DownBlock2D\",  # a regular ResNet downsampling block\n",
    "        \"DownBlock2D\",\n",
    "        \"DownBlock2D\",\n",
    "        \"DownBlock2D\",\n",
    "        \"AttnDownBlock2D\",  # a ResNet downsampling block with spatial self-attention\n",
    "        \"DownBlock2D\",\n",
    "    ),\n",
    "    up_block_types=(\n",
    "        \"UpBlock2D\",  # a regular ResNet upsampling block\n",
    "        \"AttnUpBlock2D\",  # a ResNet upsampling block with spatial self-attention\n",
    "        \"UpBlock2D\",\n",
    "        \"UpBlock2D\",\n",
    "        \"UpBlock2D\",\n",
    "        \"UpBlock2D\",\n",
    "    ),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input shape: torch.Size([1, 3, 128, 128])\n",
      "Output shape: torch.Size([1, 3, 128, 128])\n"
     ]
    }
   ],
   "source": [
    "sample_image = dataset[0][\"images\"].unsqueeze(0)\n",
    "print(\"Input shape:\", sample_image.shape)\n",
    "\n",
    "print(\"Output shape:\", model(sample_image, timestep=0).sample.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/jpeg": "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",
      "image/png": "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",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=128x128>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Scheduler\n",
    "import torch\n",
    "from PIL import Image\n",
    "from diffusers import DDPMScheduler\n",
    "\n",
    "noise_scheduler = DDPMScheduler(num_train_timesteps=1000)\n",
    "noise = torch.randn(sample_image.shape)\n",
    "timesteps = torch.LongTensor([50])\n",
    "noisy_image = noise_scheduler.add_noise(sample_image, noise, timesteps)\n",
    "\n",
    "Image.fromarray(((noisy_image.permute(0, 2, 3, 1) + 1.0) * 127.5).type(torch.uint8).numpy()[0])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.nn.functional as F\n",
    "\n",
    "noise_pred = model(noisy_image, timesteps).sample\n",
    "loss = F.mse_loss(noise_pred, noise)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Train the model\n",
    "from diffusers.optimization import get_cosine_schedule_with_warmup\n",
    "\n",
    "optimizer = torch.optim.AdamW(model.parameters(), lr=config.learning_rate)\n",
    "lr_scheduler = get_cosine_schedule_with_warmup(\n",
    "    optimizer=optimizer,\n",
    "    num_warmup_steps=config.lr_warmup_steps,\n",
    "    num_training_steps=(len(train_dataloader) * config.num_epochs),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Evaluate\n",
    "from diffusers import DDPMPipeline\n",
    "from diffusers.utils import make_image_grid\n",
    "import os\n",
    "\n",
    "def evaluate(config, epoch, pipeline):\n",
    "    # Sample some images from random noise (this is the backward diffusion process).\n",
    "    # The default pipeline output type is `List[PIL.Image]`\n",
    "    images = pipeline(\n",
    "        batch_size=config.eval_batch_size,\n",
    "        generator=torch.Generator(device='cpu').manual_seed(config.seed), # Use a separate torch generator to avoid rewinding the random state of the main training loop\n",
    "    ).images\n",
    "\n",
    "    # Make a grid out of the images\n",
    "    image_grid = make_image_grid(images, rows=4, cols=4)\n",
    "\n",
    "    # Save the images\n",
    "    test_dir = os.path.join(config.output_dir, \"samples\")\n",
    "    os.makedirs(test_dir, exist_ok=True)\n",
    "    image_grid.save(f\"{test_dir}/{epoch:04d}.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "#!pip install accelerate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from accelerate import Accelerator\n",
    "from huggingface_hub import create_repo, upload_folder\n",
    "from tqdm.auto import tqdm\n",
    "from pathlib import Path\n",
    "import os\n",
    "\n",
    "models_dir = os.path.join(config.output_dir, \"models\")\n",
    "os.makedirs(models_dir, exist_ok=True)\n",
    "\n",
    "def train_loop(config, model, noise_scheduler, optimizer, train_dataloader, lr_scheduler):\n",
    "    # Initialize accelerator and tensorboard logging\n",
    "    accelerator = Accelerator(\n",
    "        mixed_precision=config.mixed_precision,\n",
    "        gradient_accumulation_steps=config.gradient_accumulation_steps,\n",
    "        log_with=\"tensorboard\",\n",
    "        project_dir=os.path.join(config.output_dir, \"logs\"),\n",
    "    )\n",
    "    if accelerator.is_main_process:\n",
    "        if config.output_dir is not None:\n",
    "            os.makedirs(config.output_dir, exist_ok=True)\n",
    "        if config.push_to_hub:\n",
    "            repo_id = create_repo(\n",
    "                repo_id=config.hub_model_id or Path(config.output_dir).name, exist_ok=True\n",
    "            ).repo_id\n",
    "        accelerator.init_trackers(\"train_example\")\n",
    "\n",
    "    # Prepare everything\n",
    "    # There is no specific order to remember, you just need to unpack the\n",
    "    # objects in the same order you gave them to the prepare method.\n",
    "    model, optimizer, train_dataloader, lr_scheduler = accelerator.prepare(\n",
    "        model, optimizer, train_dataloader, lr_scheduler\n",
    "    )\n",
    "\n",
    "    global_step = 0\n",
    "\n",
    "    # Now you train the model\n",
    "    for epoch in range(config.num_epochs):\n",
    "        progress_bar = tqdm(total=len(train_dataloader), disable=not accelerator.is_local_main_process)\n",
    "        progress_bar.set_description(f\"Epoch {epoch}\")\n",
    "\n",
    "        for step, batch in enumerate(train_dataloader):\n",
    "            clean_images = batch[\"images\"]\n",
    "            # Sample noise to add to the images\n",
    "            noise = torch.randn(clean_images.shape, device=clean_images.device)\n",
    "            bs = clean_images.shape[0]\n",
    "\n",
    "            # Sample a random timestep for each image\n",
    "            timesteps = torch.randint(\n",
    "                0, noise_scheduler.config.num_train_timesteps, (bs,), device=clean_images.device,\n",
    "                dtype=torch.int64\n",
    "            )\n",
    "\n",
    "            # Add noise to the clean images according to the noise magnitude at each timestep\n",
    "            # (this is the forward diffusion process)\n",
    "            noisy_images = noise_scheduler.add_noise(clean_images, noise, timesteps)\n",
    "\n",
    "            with accelerator.accumulate(model):\n",
    "                # Predict the noise residual\n",
    "                noise_pred = model(noisy_images, timesteps, return_dict=False)[0]\n",
    "                loss = F.mse_loss(noise_pred, noise)\n",
    "                accelerator.backward(loss)\n",
    "\n",
    "                if accelerator.sync_gradients:\n",
    "                    accelerator.clip_grad_norm_(model.parameters(), 1.0)\n",
    "                optimizer.step()\n",
    "                lr_scheduler.step()\n",
    "                optimizer.zero_grad()\n",
    "\n",
    "            progress_bar.update(1)\n",
    "            logs = {\"loss\": loss.detach().item(), \"lr\": lr_scheduler.get_last_lr()[0], \"step\": global_step}\n",
    "            progress_bar.set_postfix(**logs)\n",
    "            accelerator.log(logs, step=global_step)\n",
    "            global_step += 1\n",
    "\n",
    "        # After each epoch you optionally sample some demo images with evaluate() and save the model\n",
    "        if accelerator.is_main_process:\n",
    "            pipeline = DDPMPipeline(unet=accelerator.unwrap_model(model), scheduler=noise_scheduler)\n",
    "\n",
    "            if (epoch + 1) % config.save_image_epochs == 0 or epoch == config.num_epochs - 1:\n",
    "                evaluate(config, epoch, pipeline)\n",
    "\n",
    "            if (epoch + 1) % config.save_model_epochs == 0 or epoch == config.num_epochs - 1:\n",
    "                if config.push_to_hub:\n",
    "                    upload_folder(\n",
    "                        repo_id=repo_id,\n",
    "                        folder_path=config.output_dir,\n",
    "                        commit_message=f\"Epoch {epoch}\",\n",
    "                        ignore_patterns=[\"step_*\", \"epoch_*\"],\n",
    "                    )\n",
    "                else:\n",
    "                    pipeline.save_pretrained(f\"{models_dir}/{epoch:04d}.pt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Launching training on one GPU.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/antonia/DIS2024/.venv/lib/python3.10/site-packages/accelerate/accelerator.py:406: UserWarning: `log_with=tensorboard` was passed but no supported trackers are currently installed.\n",
      "  warnings.warn(f\"`log_with={log_with}` was passed but no supported trackers are currently installed.\")\n",
      "/home/antonia/DIS2024/.venv/lib/python3.10/site-packages/accelerate/accelerator.py:488: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.\n",
      "  self.scaler = torch.cuda.amp.GradScaler(**kwargs)\n",
      "Epoch 0: 100%|██████████| 449/449 [02:44<00:00,  2.73it/s, loss=0.00795, lr=8.98e-5, step=448]\n",
      "Epoch 1: 100%|██████████| 449/449 [02:43<00:00,  2.74it/s, loss=0.00324, lr=0.0001, step=897]\n",
      "Epoch 2: 100%|██████████| 449/449 [02:41<00:00,  2.79it/s, loss=0.00148, lr=9.99e-5, step=1346]\n",
      "Epoch 3: 100%|██████████| 449/449 [02:44<00:00,  2.73it/s, loss=0.00653, lr=9.98e-5, step=1795]\n",
      "Epoch 4: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.0018, lr=9.96e-5, step=2244] \n",
      "Epoch 5: 100%|██████████| 449/449 [02:42<00:00,  2.77it/s, loss=0.00139, lr=9.94e-5, step=2693]\n",
      "Epoch 6: 100%|██████████| 449/449 [02:41<00:00,  2.79it/s, loss=0.00443, lr=9.91e-5, step=3142]\n",
      "Epoch 7: 100%|██████████| 449/449 [02:41<00:00,  2.77it/s, loss=0.00087, lr=9.88e-5, step=3591]\n",
      "Epoch 8: 100%|██████████| 449/449 [02:41<00:00,  2.78it/s, loss=0.001, lr=9.84e-5, step=4040]  \n",
      "100%|██████████| 1000/1000 [01:46<00:00,  9.40it/s]\n",
      "Epoch 9: 100%|██████████| 449/449 [04:29<00:00,  1.67it/s, loss=0.00287, lr=9.8e-5, step=4489]\n",
      "Epoch 10: 100%|██████████| 449/449 [02:44<00:00,  2.72it/s, loss=0.0274, lr=9.76e-5, step=4938] \n",
      "Epoch 11: 100%|██████████| 449/449 [02:44<00:00,  2.73it/s, loss=0.177, lr=9.7e-5, step=5387]\n",
      "Epoch 12: 100%|██████████| 449/449 [02:43<00:00,  2.74it/s, loss=0.000292, lr=9.65e-5, step=5836]\n",
      "Epoch 13: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.00038, lr=9.59e-5, step=6285]\n",
      "Epoch 14: 100%|██████████| 449/449 [02:41<00:00,  2.78it/s, loss=0.000258, lr=9.52e-5, step=6734]\n",
      "Epoch 15: 100%|██████████| 449/449 [02:41<00:00,  2.77it/s, loss=0.00142, lr=9.45e-5, step=7183]\n",
      "Epoch 16: 100%|██████████| 449/449 [02:45<00:00,  2.71it/s, loss=0.000419, lr=9.38e-5, step=7632]\n",
      "Epoch 17: 100%|██████████| 449/449 [02:45<00:00,  2.71it/s, loss=0.000565, lr=9.3e-5, step=8081]\n",
      "Epoch 18: 100%|██████████| 449/449 [02:41<00:00,  2.78it/s, loss=0.0288, lr=9.21e-5, step=8530] \n",
      "100%|██████████| 1000/1000 [01:46<00:00,  9.39it/s]\n",
      "Epoch 19: 100%|██████████| 449/449 [04:29<00:00,  1.67it/s, loss=0.00401, lr=9.13e-5, step=8979]\n",
      "Epoch 20: 100%|██████████| 449/449 [02:43<00:00,  2.74it/s, loss=0.0336, lr=9.03e-5, step=9428] \n",
      "Epoch 21: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.0373, lr=8.94e-5, step=9877]\n",
      "Epoch 22: 100%|██████████| 449/449 [02:41<00:00,  2.79it/s, loss=0.141, lr=8.84e-5, step=10326]  \n",
      "Epoch 23: 100%|██████████| 449/449 [02:44<00:00,  2.73it/s, loss=0.000379, lr=8.74e-5, step=10775]\n",
      "Epoch 24: 100%|██████████| 449/449 [02:45<00:00,  2.72it/s, loss=0.000491, lr=8.63e-5, step=11224]\n",
      "Epoch 25: 100%|██████████| 449/449 [02:45<00:00,  2.71it/s, loss=0.0101, lr=8.52e-5, step=11673]\n",
      "Epoch 26: 100%|██████████| 449/449 [02:43<00:00,  2.75it/s, loss=0.0343, lr=8.4e-5, step=12122]  \n",
      "Epoch 27: 100%|██████████| 449/449 [02:44<00:00,  2.72it/s, loss=0.000219, lr=8.28e-5, step=12571]\n",
      "Epoch 28: 100%|██████████| 449/449 [02:45<00:00,  2.72it/s, loss=0.00106, lr=8.16e-5, step=13020]\n",
      "100%|██████████| 1000/1000 [01:46<00:00,  9.39it/s]\n",
      "Epoch 29: 100%|██████████| 449/449 [04:33<00:00,  1.64it/s, loss=0.0251, lr=8.04e-5, step=13469]\n",
      "Epoch 30: 100%|██████████| 449/449 [02:44<00:00,  2.72it/s, loss=0.000565, lr=7.91e-5, step=13918]\n",
      "Epoch 31: 100%|██████████| 449/449 [02:43<00:00,  2.74it/s, loss=0.0057, lr=7.78e-5, step=14367]\n",
      "Epoch 32: 100%|██████████| 449/449 [02:43<00:00,  2.75it/s, loss=0.000485, lr=7.65e-5, step=14816]\n",
      "Epoch 33: 100%|██████████| 449/449 [02:41<00:00,  2.78it/s, loss=0.0206, lr=7.51e-5, step=15265]\n",
      "Epoch 34: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.000194, lr=7.37e-5, step=15714]\n",
      "Epoch 35: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.000385, lr=7.23e-5, step=16163]\n",
      "Epoch 36: 100%|██████████| 449/449 [02:41<00:00,  2.77it/s, loss=0.0237, lr=7.09e-5, step=16612] \n",
      "Epoch 37: 100%|██████████| 449/449 [02:41<00:00,  2.78it/s, loss=0.00726, lr=6.94e-5, step=17061]\n",
      "Epoch 38: 100%|██████████| 449/449 [02:43<00:00,  2.75it/s, loss=0.0129, lr=6.8e-5, step=17510]  \n",
      "100%|██████████| 1000/1000 [01:46<00:00,  9.40it/s]\n",
      "Epoch 39: 100%|██████████| 449/449 [04:29<00:00,  1.67it/s, loss=0.00106, lr=6.65e-5, step=17959]\n",
      "Epoch 40: 100%|██████████| 449/449 [02:41<00:00,  2.79it/s, loss=0.000287, lr=6.5e-5, step=18408]\n",
      "Epoch 41: 100%|██████████| 449/449 [02:41<00:00,  2.77it/s, loss=0.00114, lr=6.34e-5, step=18857]\n",
      "Epoch 42: 100%|██████████| 449/449 [02:41<00:00,  2.78it/s, loss=0.00592, lr=6.19e-5, step=19306]\n",
      "Epoch 43: 100%|██████████| 449/449 [02:42<00:00,  2.77it/s, loss=0.0296, lr=6.04e-5, step=19755]\n",
      "Epoch 44: 100%|██████████| 449/449 [02:42<00:00,  2.77it/s, loss=0.00966, lr=5.88e-5, step=20204]\n",
      "Epoch 45: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.00131, lr=5.72e-5, step=20653]\n",
      "Epoch 46: 100%|██████████| 449/449 [02:41<00:00,  2.78it/s, loss=0.000553, lr=5.57e-5, step=21102]\n",
      "Epoch 47: 100%|██████████| 449/449 [02:41<00:00,  2.79it/s, loss=0.00728, lr=5.41e-5, step=21551]\n",
      "Epoch 48: 100%|██████████| 449/449 [02:41<00:00,  2.79it/s, loss=0.0203, lr=5.25e-5, step=22000] \n",
      "100%|██████████| 1000/1000 [01:46<00:00,  9.42it/s]\n",
      "Epoch 49: 100%|██████████| 449/449 [04:28<00:00,  1.67it/s, loss=0.000304, lr=5.09e-5, step=22449]\n",
      "Epoch 50: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.0103, lr=4.93e-5, step=22898] \n",
      "Epoch 51: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.00618, lr=4.77e-5, step=23347]\n",
      "Epoch 52: 100%|██████████| 449/449 [02:44<00:00,  2.73it/s, loss=0.00354, lr=4.61e-5, step=23796]\n",
      "Epoch 53: 100%|██████████| 449/449 [02:43<00:00,  2.74it/s, loss=0.000164, lr=4.46e-5, step=24245]\n",
      "Epoch 54: 100%|██████████| 449/449 [02:44<00:00,  2.73it/s, loss=0.0365, lr=4.3e-5, step=24694]  \n",
      "Epoch 55: 100%|██████████| 449/449 [02:44<00:00,  2.74it/s, loss=0.000133, lr=4.14e-5, step=25143]\n",
      "Epoch 56: 100%|██████████| 449/449 [02:40<00:00,  2.80it/s, loss=0.00696, lr=3.99e-5, step=25592]\n",
      "Epoch 57: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.0105, lr=3.83e-5, step=26041]\n",
      "Epoch 58: 100%|██████████| 449/449 [02:43<00:00,  2.75it/s, loss=0.00579, lr=3.68e-5, step=26490]\n",
      "100%|██████████| 1000/1000 [01:46<00:00,  9.41it/s]\n",
      "Epoch 59: 100%|██████████| 449/449 [04:29<00:00,  1.67it/s, loss=0.0128, lr=3.52e-5, step=26939]\n",
      "Epoch 60: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.00648, lr=3.37e-5, step=27388]\n",
      "Epoch 61: 100%|██████████| 449/449 [02:43<00:00,  2.74it/s, loss=0.00489, lr=3.22e-5, step=27837]\n",
      "Epoch 62: 100%|██████████| 449/449 [02:41<00:00,  2.78it/s, loss=0.0578, lr=3.08e-5, step=28286] \n",
      "Epoch 63: 100%|██████████| 449/449 [02:41<00:00,  2.77it/s, loss=0.00223, lr=2.93e-5, step=28735]\n",
      "Epoch 64: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.0053, lr=2.79e-5, step=29184] \n",
      "Epoch 65: 100%|██████████| 449/449 [02:42<00:00,  2.77it/s, loss=0.0345, lr=2.65e-5, step=29633]\n",
      "Epoch 66: 100%|██████████| 449/449 [02:44<00:00,  2.74it/s, loss=0.0063, lr=2.51e-5, step=30082] \n",
      "Epoch 67: 100%|██████████| 449/449 [02:45<00:00,  2.72it/s, loss=0.000469, lr=2.37e-5, step=30531]\n",
      "Epoch 68: 100%|██████████| 449/449 [02:43<00:00,  2.74it/s, loss=0.0453, lr=2.24e-5, step=30980] \n",
      "100%|██████████| 1000/1000 [01:46<00:00,  9.40it/s]\n",
      "Epoch 69: 100%|██████████| 449/449 [04:32<00:00,  1.64it/s, loss=0.167, lr=2.11e-5, step=31429]\n",
      "Epoch 70: 100%|██████████| 449/449 [02:45<00:00,  2.72it/s, loss=0.000202, lr=1.98e-5, step=31878]\n",
      "Epoch 71: 100%|██████████| 449/449 [02:43<00:00,  2.74it/s, loss=0.00115, lr=1.85e-5, step=32327]\n",
      "Epoch 72: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.0958, lr=1.73e-5, step=32776] \n",
      "Epoch 73: 100%|██████████| 449/449 [02:41<00:00,  2.78it/s, loss=0.0094, lr=1.61e-5, step=33225]\n",
      "Epoch 74: 100%|██████████| 449/449 [02:41<00:00,  2.79it/s, loss=0.00887, lr=1.5e-5, step=33674] \n",
      "Epoch 75: 100%|██████████| 449/449 [02:41<00:00,  2.78it/s, loss=0.000104, lr=1.39e-5, step=34123]\n",
      "Epoch 76: 100%|██████████| 449/449 [02:43<00:00,  2.75it/s, loss=0.00513, lr=1.28e-5, step=34572]\n",
      "Epoch 77: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.00499, lr=1.17e-5, step=35021]\n",
      "Epoch 78: 100%|██████████| 449/449 [02:40<00:00,  2.79it/s, loss=0.0219, lr=1.07e-5, step=35470] \n",
      "100%|██████████| 1000/1000 [01:46<00:00,  9.39it/s]\n",
      "Epoch 79: 100%|██████████| 449/449 [04:30<00:00,  1.66it/s, loss=0.000138, lr=9.78e-6, step=35919]\n",
      "Epoch 80: 100%|██████████| 449/449 [02:42<00:00,  2.76it/s, loss=0.00277, lr=8.86e-6, step=36368]\n",
      "Epoch 81: 100%|██████████| 449/449 [02:41<00:00,  2.78it/s, loss=8.59e-5, lr=7.98e-6, step=36817]\n",
      "Epoch 82: 100%|██████████| 449/449 [02:41<00:00,  2.78it/s, loss=0.0582, lr=7.14e-6, step=37266] \n",
      "Epoch 83: 100%|██████████| 449/449 [02:44<00:00,  2.74it/s, loss=0.00134, lr=6.34e-6, step=37715]\n",
      "Epoch 84: 100%|██████████| 449/449 [02:43<00:00,  2.75it/s, loss=0.000167, lr=5.59e-6, step=38164]\n",
      "Epoch 85: 100%|██████████| 449/449 [02:43<00:00,  2.75it/s, loss=0.0525, lr=4.88e-6, step=38613]\n",
      "Epoch 86: 100%|██████████| 449/449 [02:44<00:00,  2.74it/s, loss=0.00825, lr=4.22e-6, step=39062]\n",
      "Epoch 87: 100%|██████████| 449/449 [02:43<00:00,  2.74it/s, loss=0.000163, lr=3.61e-6, step=39511]\n",
      "Epoch 88: 100%|██████████| 449/449 [02:40<00:00,  2.79it/s, loss=0.0259, lr=3.04e-6, step=4e+4]  \n",
      "100%|██████████| 1000/1000 [01:46<00:00,  9.38it/s]\n",
      "Epoch 89: 100%|██████████| 449/449 [04:29<00:00,  1.67it/s, loss=0.00015, lr=2.52e-6, step=40409]\n",
      "Epoch 90: 100%|██████████| 449/449 [02:44<00:00,  2.72it/s, loss=0.0192, lr=2.04e-6, step=40858] \n",
      "Epoch 91: 100%|██████████| 449/449 [02:42<00:00,  2.77it/s, loss=0.000873, lr=1.62e-6, step=41307]\n",
      "Epoch 92: 100%|██████████| 449/449 [02:42<00:00,  2.77it/s, loss=0.00376, lr=1.24e-6, step=41756]\n",
      "Epoch 93: 100%|██████████| 449/449 [02:45<00:00,  2.71it/s, loss=0.0675, lr=9.14e-7, step=42205]\n",
      "Epoch 94: 100%|██████████| 449/449 [02:43<00:00,  2.75it/s, loss=0.000409, lr=6.37e-7, step=42654]\n",
      "Epoch 95: 100%|██████████| 449/449 [02:43<00:00,  2.74it/s, loss=0.0493, lr=4.09e-7, step=43103]\n",
      "Epoch 96: 100%|██████████| 449/449 [02:44<00:00,  2.74it/s, loss=0.0137, lr=2.31e-7, step=43552] \n",
      "Epoch 97: 100%|██████████| 449/449 [02:41<00:00,  2.78it/s, loss=0.000883, lr=1.04e-7, step=44001]\n",
      "Epoch 98: 100%|██████████| 449/449 [02:41<00:00,  2.79it/s, loss=0.000727, lr=2.69e-8, step=44450]\n",
      "100%|██████████| 1000/1000 [01:46<00:00,  9.38it/s]\n",
      "Epoch 99: 100%|██████████| 449/449 [04:31<00:00,  1.65it/s, loss=0.015, lr=2.82e-11, step=44899]\n"
     ]
    }
   ],
   "source": [
    "from accelerate import notebook_launcher\n",
    "\n",
    "args = (config, model, noise_scheduler, optimizer, train_dataloader, lr_scheduler)\n",
    "\n",
    "notebook_launcher(train_loop, args, num_processes=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Loading pipeline components...: 100%|██████████| 2/2 [00:00<00:00, 18.12it/s]\n"
     ]
    }
   ],
   "source": [
    "# Generate image with the model of choice\n",
    "from diffusers import DiffusionPipeline\n",
    "\n",
    "pipeline = DiffusionPipeline.from_pretrained(\"flowers_August_100/models/0099.pt\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "image = pipeline().images[0]  # Generate one image at a time\n",
    "image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [05:28<00:00,  3.04it/s]\n",
      " 33%|███▎      | 327/1000 [01:46<03:39,  3.06it/s]\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[28], line 9\u001b[0m\n\u001b[1;32m      7\u001b[0m \u001b[38;5;66;03m# Generate 6 images\u001b[39;00m\n\u001b[1;32m      8\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m _ \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mrange\u001b[39m(\u001b[38;5;241m6\u001b[39m):\n\u001b[0;32m----> 9\u001b[0m     image \u001b[38;5;241m=\u001b[39m \u001b[43mpipeline\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241m.\u001b[39mimages[\u001b[38;5;241m0\u001b[39m]  \u001b[38;5;66;03m# Generate one image at a time\u001b[39;00m\n\u001b[1;32m     10\u001b[0m     images\u001b[38;5;241m.\u001b[39mappend(image)\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/utils/_contextlib.py:116\u001b[0m, in \u001b[0;36mcontext_decorator.<locals>.decorate_context\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    113\u001b[0m \u001b[38;5;129m@functools\u001b[39m\u001b[38;5;241m.\u001b[39mwraps(func)\n\u001b[1;32m    114\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mdecorate_context\u001b[39m(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs):\n\u001b[1;32m    115\u001b[0m     \u001b[38;5;28;01mwith\u001b[39;00m ctx_factory():\n\u001b[0;32m--> 116\u001b[0m         \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/diffusers/pipelines/ddpm/pipeline_ddpm.py:114\u001b[0m, in \u001b[0;36mDDPMPipeline.__call__\u001b[0;34m(self, batch_size, generator, num_inference_steps, output_type, return_dict)\u001b[0m\n\u001b[1;32m    110\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mscheduler\u001b[38;5;241m.\u001b[39mset_timesteps(num_inference_steps)\n\u001b[1;32m    112\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m t \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mprogress_bar(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mscheduler\u001b[38;5;241m.\u001b[39mtimesteps):\n\u001b[1;32m    113\u001b[0m     \u001b[38;5;66;03m# 1. predict noise model_output\u001b[39;00m\n\u001b[0;32m--> 114\u001b[0m     model_output \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43munet\u001b[49m\u001b[43m(\u001b[49m\u001b[43mimage\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mt\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241m.\u001b[39msample\n\u001b[1;32m    116\u001b[0m     \u001b[38;5;66;03m# 2. compute previous image: x_t -> x_t-1\u001b[39;00m\n\u001b[1;32m    117\u001b[0m     image \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mscheduler\u001b[38;5;241m.\u001b[39mstep(model_output, t, image, generator\u001b[38;5;241m=\u001b[39mgenerator)\u001b[38;5;241m.\u001b[39mprev_sample\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1553\u001b[0m, in \u001b[0;36mModule._wrapped_call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1551\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_compiled_call_impl(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)  \u001b[38;5;66;03m# type: ignore[misc]\u001b[39;00m\n\u001b[1;32m   1552\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m-> 1553\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call_impl\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1562\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1557\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1558\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1559\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks\n\u001b[1;32m   1560\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1561\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1562\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1564\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m   1565\u001b[0m     result \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/diffusers/models/unets/unet_2d.py:313\u001b[0m, in \u001b[0;36mUNet2DModel.forward\u001b[0;34m(self, sample, timestep, class_labels, return_dict)\u001b[0m\n\u001b[1;32m    309\u001b[0m         sample, res_samples, skip_sample \u001b[38;5;241m=\u001b[39m downsample_block(\n\u001b[1;32m    310\u001b[0m             hidden_states\u001b[38;5;241m=\u001b[39msample, temb\u001b[38;5;241m=\u001b[39memb, skip_sample\u001b[38;5;241m=\u001b[39mskip_sample\n\u001b[1;32m    311\u001b[0m         )\n\u001b[1;32m    312\u001b[0m     \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m--> 313\u001b[0m         sample, res_samples \u001b[38;5;241m=\u001b[39m \u001b[43mdownsample_block\u001b[49m\u001b[43m(\u001b[49m\u001b[43mhidden_states\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43msample\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtemb\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43memb\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    315\u001b[0m     down_block_res_samples \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m res_samples\n\u001b[1;32m    317\u001b[0m \u001b[38;5;66;03m# 4. mid\u001b[39;00m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1553\u001b[0m, in \u001b[0;36mModule._wrapped_call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1551\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_compiled_call_impl(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)  \u001b[38;5;66;03m# type: ignore[misc]\u001b[39;00m\n\u001b[1;32m   1552\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m-> 1553\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call_impl\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1562\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1557\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1558\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1559\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks\n\u001b[1;32m   1560\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1561\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1562\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1564\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m   1565\u001b[0m     result \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/diffusers/models/unets/unet_2d_blocks.py:1391\u001b[0m, in \u001b[0;36mDownBlock2D.forward\u001b[0;34m(self, hidden_states, temb, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1387\u001b[0m             hidden_states \u001b[38;5;241m=\u001b[39m torch\u001b[38;5;241m.\u001b[39mutils\u001b[38;5;241m.\u001b[39mcheckpoint\u001b[38;5;241m.\u001b[39mcheckpoint(\n\u001b[1;32m   1388\u001b[0m                 create_custom_forward(resnet), hidden_states, temb\n\u001b[1;32m   1389\u001b[0m             )\n\u001b[1;32m   1390\u001b[0m     \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m-> 1391\u001b[0m         hidden_states \u001b[38;5;241m=\u001b[39m \u001b[43mresnet\u001b[49m\u001b[43m(\u001b[49m\u001b[43mhidden_states\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtemb\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1393\u001b[0m     output_states \u001b[38;5;241m=\u001b[39m output_states \u001b[38;5;241m+\u001b[39m (hidden_states,)\n\u001b[1;32m   1395\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdownsamplers \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1553\u001b[0m, in \u001b[0;36mModule._wrapped_call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1551\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_compiled_call_impl(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)  \u001b[38;5;66;03m# type: ignore[misc]\u001b[39;00m\n\u001b[1;32m   1552\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m-> 1553\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call_impl\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1562\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1557\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1558\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1559\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks\n\u001b[1;32m   1560\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1561\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1562\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1564\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m   1565\u001b[0m     result \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/diffusers/models/resnet.py:366\u001b[0m, in \u001b[0;36mResnetBlock2D.forward\u001b[0;34m(self, input_tensor, temb, *args, **kwargs)\u001b[0m\n\u001b[1;32m    363\u001b[0m hidden_states \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mnonlinearity(hidden_states)\n\u001b[1;32m    365\u001b[0m hidden_states \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdropout(hidden_states)\n\u001b[0;32m--> 366\u001b[0m hidden_states \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mconv2\u001b[49m\u001b[43m(\u001b[49m\u001b[43mhidden_states\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    368\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mconv_shortcut \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m    369\u001b[0m     input_tensor \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mconv_shortcut(input_tensor)\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1553\u001b[0m, in \u001b[0;36mModule._wrapped_call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1551\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_compiled_call_impl(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)  \u001b[38;5;66;03m# type: ignore[misc]\u001b[39;00m\n\u001b[1;32m   1552\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m-> 1553\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call_impl\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1562\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1557\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1558\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1559\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks\n\u001b[1;32m   1560\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1561\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1562\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1564\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m   1565\u001b[0m     result \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/conv.py:458\u001b[0m, in \u001b[0;36mConv2d.forward\u001b[0;34m(self, input)\u001b[0m\n\u001b[1;32m    457\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mforward\u001b[39m(\u001b[38;5;28mself\u001b[39m, \u001b[38;5;28minput\u001b[39m: Tensor) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m Tensor:\n\u001b[0;32m--> 458\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_conv_forward\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mweight\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbias\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/conv.py:454\u001b[0m, in \u001b[0;36mConv2d._conv_forward\u001b[0;34m(self, input, weight, bias)\u001b[0m\n\u001b[1;32m    450\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mpadding_mode \u001b[38;5;241m!=\u001b[39m \u001b[38;5;124m'\u001b[39m\u001b[38;5;124mzeros\u001b[39m\u001b[38;5;124m'\u001b[39m:\n\u001b[1;32m    451\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m F\u001b[38;5;241m.\u001b[39mconv2d(F\u001b[38;5;241m.\u001b[39mpad(\u001b[38;5;28minput\u001b[39m, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_reversed_padding_repeated_twice, mode\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mpadding_mode),\n\u001b[1;32m    452\u001b[0m                     weight, bias, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mstride,\n\u001b[1;32m    453\u001b[0m                     _pair(\u001b[38;5;241m0\u001b[39m), \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdilation, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mgroups)\n\u001b[0;32m--> 454\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mF\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mconv2d\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mweight\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mbias\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstride\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    455\u001b[0m \u001b[43m                \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mpadding\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdilation\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mgroups\u001b[49m\u001b[43m)\u001b[49m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "from PIL import Image\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Create a list to hold the images\n",
    "images = []\n",
    "\n",
    "# Generate 6 images\n",
    "for _ in range(6):\n",
    "    image = pipeline().images[0]  # Generate one image at a time\n",
    "    images.append(image)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1800x400 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create a figure with 6 subplots in one row\n",
    "fig, axs = plt.subplots(1, 6, figsize=(18, 4))  # 1 row, 6 columns\n",
    "\n",
    "# Loop through each image and subplot axis\n",
    "for i, image in enumerate(images):\n",
    "    axs[i].imshow(image)\n",
    "    axs[i].axis('off')\n",
    "    axs[i].set_title(f\"Image {i+1}\")\n",
    "\n",
    "# Show the plot\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  2%|▏         | 16/1000 [00:05<05:27,  3.00it/s]\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[22], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m image \u001b[38;5;241m=\u001b[39m \u001b[43mpipeline\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241m.\u001b[39mimages[\u001b[38;5;241m0\u001b[39m]\n\u001b[1;32m      3\u001b[0m image\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/utils/_contextlib.py:116\u001b[0m, in \u001b[0;36mcontext_decorator.<locals>.decorate_context\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    113\u001b[0m \u001b[38;5;129m@functools\u001b[39m\u001b[38;5;241m.\u001b[39mwraps(func)\n\u001b[1;32m    114\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mdecorate_context\u001b[39m(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs):\n\u001b[1;32m    115\u001b[0m     \u001b[38;5;28;01mwith\u001b[39;00m ctx_factory():\n\u001b[0;32m--> 116\u001b[0m         \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/diffusers/pipelines/ddpm/pipeline_ddpm.py:114\u001b[0m, in \u001b[0;36mDDPMPipeline.__call__\u001b[0;34m(self, batch_size, generator, num_inference_steps, output_type, return_dict)\u001b[0m\n\u001b[1;32m    110\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mscheduler\u001b[38;5;241m.\u001b[39mset_timesteps(num_inference_steps)\n\u001b[1;32m    112\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m t \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mprogress_bar(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mscheduler\u001b[38;5;241m.\u001b[39mtimesteps):\n\u001b[1;32m    113\u001b[0m     \u001b[38;5;66;03m# 1. predict noise model_output\u001b[39;00m\n\u001b[0;32m--> 114\u001b[0m     model_output \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43munet\u001b[49m\u001b[43m(\u001b[49m\u001b[43mimage\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mt\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241m.\u001b[39msample\n\u001b[1;32m    116\u001b[0m     \u001b[38;5;66;03m# 2. compute previous image: x_t -> x_t-1\u001b[39;00m\n\u001b[1;32m    117\u001b[0m     image \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mscheduler\u001b[38;5;241m.\u001b[39mstep(model_output, t, image, generator\u001b[38;5;241m=\u001b[39mgenerator)\u001b[38;5;241m.\u001b[39mprev_sample\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1553\u001b[0m, in \u001b[0;36mModule._wrapped_call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1551\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_compiled_call_impl(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)  \u001b[38;5;66;03m# type: ignore[misc]\u001b[39;00m\n\u001b[1;32m   1552\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m-> 1553\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call_impl\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1562\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1557\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1558\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1559\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks\n\u001b[1;32m   1560\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1561\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1562\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1564\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m   1565\u001b[0m     result \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/diffusers/models/unets/unet_2d.py:313\u001b[0m, in \u001b[0;36mUNet2DModel.forward\u001b[0;34m(self, sample, timestep, class_labels, return_dict)\u001b[0m\n\u001b[1;32m    309\u001b[0m         sample, res_samples, skip_sample \u001b[38;5;241m=\u001b[39m downsample_block(\n\u001b[1;32m    310\u001b[0m             hidden_states\u001b[38;5;241m=\u001b[39msample, temb\u001b[38;5;241m=\u001b[39memb, skip_sample\u001b[38;5;241m=\u001b[39mskip_sample\n\u001b[1;32m    311\u001b[0m         )\n\u001b[1;32m    312\u001b[0m     \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m--> 313\u001b[0m         sample, res_samples \u001b[38;5;241m=\u001b[39m \u001b[43mdownsample_block\u001b[49m\u001b[43m(\u001b[49m\u001b[43mhidden_states\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43msample\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtemb\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43memb\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    315\u001b[0m     down_block_res_samples \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m res_samples\n\u001b[1;32m    317\u001b[0m \u001b[38;5;66;03m# 4. mid\u001b[39;00m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1553\u001b[0m, in \u001b[0;36mModule._wrapped_call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1551\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_compiled_call_impl(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)  \u001b[38;5;66;03m# type: ignore[misc]\u001b[39;00m\n\u001b[1;32m   1552\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m-> 1553\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call_impl\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1562\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1557\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1558\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1559\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks\n\u001b[1;32m   1560\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1561\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1562\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1564\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m   1565\u001b[0m     result \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/diffusers/models/unets/unet_2d_blocks.py:1397\u001b[0m, in \u001b[0;36mDownBlock2D.forward\u001b[0;34m(self, hidden_states, temb, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1395\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdownsamplers \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m   1396\u001b[0m     \u001b[38;5;28;01mfor\u001b[39;00m downsampler \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdownsamplers:\n\u001b[0;32m-> 1397\u001b[0m         hidden_states \u001b[38;5;241m=\u001b[39m \u001b[43mdownsampler\u001b[49m\u001b[43m(\u001b[49m\u001b[43mhidden_states\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1399\u001b[0m     output_states \u001b[38;5;241m=\u001b[39m output_states \u001b[38;5;241m+\u001b[39m (hidden_states,)\n\u001b[1;32m   1401\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m hidden_states, output_states\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1553\u001b[0m, in \u001b[0;36mModule._wrapped_call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1551\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_compiled_call_impl(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)  \u001b[38;5;66;03m# type: ignore[misc]\u001b[39;00m\n\u001b[1;32m   1552\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m-> 1553\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call_impl\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1562\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1557\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1558\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1559\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks\n\u001b[1;32m   1560\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1561\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1562\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1564\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m   1565\u001b[0m     result \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/diffusers/models/downsampling.py:147\u001b[0m, in \u001b[0;36mDownsample2D.forward\u001b[0;34m(self, hidden_states, *args, **kwargs)\u001b[0m\n\u001b[1;32m    143\u001b[0m     hidden_states \u001b[38;5;241m=\u001b[39m F\u001b[38;5;241m.\u001b[39mpad(hidden_states, pad, mode\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mconstant\u001b[39m\u001b[38;5;124m\"\u001b[39m, value\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m0\u001b[39m)\n\u001b[1;32m    145\u001b[0m \u001b[38;5;28;01massert\u001b[39;00m hidden_states\u001b[38;5;241m.\u001b[39mshape[\u001b[38;5;241m1\u001b[39m] \u001b[38;5;241m==\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mchannels\n\u001b[0;32m--> 147\u001b[0m hidden_states \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mconv\u001b[49m\u001b[43m(\u001b[49m\u001b[43mhidden_states\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    149\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m hidden_states\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1553\u001b[0m, in \u001b[0;36mModule._wrapped_call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1551\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_compiled_call_impl(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)  \u001b[38;5;66;03m# type: ignore[misc]\u001b[39;00m\n\u001b[1;32m   1552\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m-> 1553\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call_impl\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1562\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1557\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1558\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1559\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks\n\u001b[1;32m   1560\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_backward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1561\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1562\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1564\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m   1565\u001b[0m     result \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/conv.py:458\u001b[0m, in \u001b[0;36mConv2d.forward\u001b[0;34m(self, input)\u001b[0m\n\u001b[1;32m    457\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mforward\u001b[39m(\u001b[38;5;28mself\u001b[39m, \u001b[38;5;28minput\u001b[39m: Tensor) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m Tensor:\n\u001b[0;32m--> 458\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_conv_forward\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mweight\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbias\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/nn/modules/conv.py:454\u001b[0m, in \u001b[0;36mConv2d._conv_forward\u001b[0;34m(self, input, weight, bias)\u001b[0m\n\u001b[1;32m    450\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mpadding_mode \u001b[38;5;241m!=\u001b[39m \u001b[38;5;124m'\u001b[39m\u001b[38;5;124mzeros\u001b[39m\u001b[38;5;124m'\u001b[39m:\n\u001b[1;32m    451\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m F\u001b[38;5;241m.\u001b[39mconv2d(F\u001b[38;5;241m.\u001b[39mpad(\u001b[38;5;28minput\u001b[39m, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_reversed_padding_repeated_twice, mode\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mpadding_mode),\n\u001b[1;32m    452\u001b[0m                     weight, bias, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mstride,\n\u001b[1;32m    453\u001b[0m                     _pair(\u001b[38;5;241m0\u001b[39m), \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdilation, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mgroups)\n\u001b[0;32m--> 454\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mF\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mconv2d\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mweight\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mbias\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstride\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    455\u001b[0m \u001b[43m                \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mpadding\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdilation\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mgroups\u001b[49m\u001b[43m)\u001b[49m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "image = pipeline().images[0]\n",
    "\n",
    "image"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Additional training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn.functional as F\n",
    "from accelerate import Accelerator\n",
    "from tqdm.auto import tqdm\n",
    "from pathlib import Path\n",
    "import os\n",
    "from diffusers import DDPMPipeline  # Import your specific pipeline\n",
    "from torch.optim import Adam\n",
    "\n",
    "def fine_tune_train_loop(config, model, noise_scheduler, optimizer, train_dataloader, lr_scheduler, start_epoch=0):\n",
    "    # Initialize accelerator and tensorboard logging\n",
    "    accelerator = Accelerator(\n",
    "        mixed_precision=config.mixed_precision,\n",
    "        gradient_accumulation_steps=config.gradient_accumulation_steps,\n",
    "        log_with=\"tensorboard\",\n",
    "        project_dir=os.path.join(config.output_dir, \"logs\"),\n",
    "    )\n",
    "    if accelerator.is_main_process:\n",
    "        if config.output_dir is not None:\n",
    "            os.makedirs(config.output_dir, exist_ok=True)\n",
    "        if config.push_to_hub:\n",
    "            repo_id = create_repo(\n",
    "                repo_id=config.hub_model_id or Path(config.output_dir).name, exist_ok=True\n",
    "            ).repo_id\n",
    "        accelerator.init_trackers(\"train_example\")\n",
    "\n",
    "    # Prepare everything\n",
    "    model, optimizer, train_dataloader, lr_scheduler = accelerator.prepare(\n",
    "        model, optimizer, train_dataloader, lr_scheduler\n",
    "    )\n",
    "\n",
    "    global_step = 0\n",
    "\n",
    "    # Now continue training the model\n",
    "    for epoch in range(start_epoch, start_epoch + config.additional_epochs):\n",
    "        progress_bar = tqdm(total=len(train_dataloader), disable=not accelerator.is_local_main_process)\n",
    "        progress_bar.set_description(f\"Epoch {epoch}\")\n",
    "\n",
    "        for step, batch in enumerate(train_dataloader):\n",
    "            clean_images = batch[\"images\"]\n",
    "            noise = torch.randn(clean_images.shape, device=clean_images.device)\n",
    "            bs = clean_images.shape[0]\n",
    "\n",
    "            timesteps = torch.randint(\n",
    "                0, noise_scheduler.config.num_train_timesteps, (bs,), device=clean_images.device,\n",
    "                dtype=torch.int64\n",
    "            )\n",
    "\n",
    "            noisy_images = noise_scheduler.add_noise(clean_images, noise, timesteps)\n",
    "\n",
    "            with accelerator.accumulate(model):\n",
    "                noise_pred = model(noisy_images, timesteps, return_dict=False)[0]\n",
    "                loss = F.mse_loss(noise_pred, noise)\n",
    "                accelerator.backward(loss)\n",
    "\n",
    "                if accelerator.sync_gradients:\n",
    "                    accelerator.clip_grad_norm_(model.parameters(), 1.0)\n",
    "                optimizer.step()\n",
    "                lr_scheduler.step()\n",
    "                optimizer.zero_grad()\n",
    "\n",
    "            progress_bar.update(1)\n",
    "            logs = {\"loss\": loss.detach().item(), \"lr\": lr_scheduler.get_last_lr()[0], \"step\": global_step}\n",
    "            progress_bar.set_postfix(**logs)\n",
    "            accelerator.log(logs, step=global_step)\n",
    "            global_step += 1\n",
    "\n",
    "        # After each epoch you optionally sample some demo images with evaluate() and save the model\n",
    "        if accelerator.is_main_process:\n",
    "            pipeline = DDPMPipeline(unet=accelerator.unwrap_model(model), scheduler=noise_scheduler)\n",
    "\n",
    "            if (epoch + 1) % config.save_image_epochs == 0 or epoch == start_epoch + config.additional_epochs - 1:\n",
    "                evaluate(config, epoch, pipeline)\n",
    "\n",
    "            if (epoch + 1) % config.save_model_epochs == 0 or epoch == start_epoch + config.additional_epochs - 1:\n",
    "                if config.push_to_hub:\n",
    "                    upload_folder(\n",
    "                        repo_id=repo_id,\n",
    "                        folder_path=config.output_dir,\n",
    "                        commit_message=f\"Epoch {epoch}\",\n",
    "                        ignore_patterns=[\"step_*\", \"epoch_*\"],\n",
    "                    )\n",
    "                else:\n",
    "                    pipeline.save_pretrained(f\"{config.output_dir}/epoch_{epoch:04d}.pt\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Loading pipeline components...: 100%|██████████| 2/2 [00:00<00:00, 17.47it/s]\n"
     ]
    }
   ],
   "source": [
    "from accelerate import notebook_launcher\n",
    "\n",
    "# Define the additional number of epochs you want to run\n",
    "config.additional_epochs = 40  # Change this to the number of additional epochs you want\n",
    "config.output_dir = \"newAugust\"  # Ensure this is the correct path to your output directory\n",
    "\n",
    "# Load the previously saved model\n",
    "pipeline = DDPMPipeline.from_pretrained(\"flowers_August0049.pt\")\n",
    "pipeline.unet = model\n",
    "\n",
    "# Reinitialize the optimizer and scheduler\n",
    "optimizer = Adam(model.parameters(), lr=config.learning_rate)  # Use the same learning rate as before\n",
    "lr_scheduler = get_cosine_schedule_with_warmup(\n",
    "    optimizer=optimizer,\n",
    "    num_warmup_steps=config.lr_warmup_steps,\n",
    "    num_training_steps=(len(train_dataloader) * config.num_epochs),\n",
    ")\n",
    "\n",
    "# Start training from the last epoch + 1\n",
    "start_epoch = 60  # Assuming the previous training was completed at epoch 50\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Launching training on one GPU.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\u001b[A\n",
      "\u001b[A"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "Epoch 50:   6%|▋         | 29/449 [01:52<27:05,  3.87s/it, loss=0.954, lr=5.8e-6, step=28]\n",
      "\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "Epoch 60: 100%|██████████| 449/449 [02:49<00:00,  2.65it/s, loss=0.00758, lr=9.36e-5, step=448]\n",
      "Epoch 61: 100%|██████████| 449/449 [02:48<00:00,  3.40it/s, loss=0.0091, lr=9.99e-5, step=897] \n",
      "Epoch 61: 100%|██████████| 449/449 [02:48<00:00,  2.67it/s, loss=0.0091, lr=9.99e-5, step=897]\n",
      "\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "Epoch 62: 100%|██████████| 449/449 [02:51<00:00,  2.62it/s, loss=0.00134, lr=9.96e-5, step=1346]\n",
      "Epoch 63: 100%|██████████| 449/449 [02:49<00:00,  3.58it/s, loss=0.00172, lr=9.91e-5, step=1795]\n",
      "Epoch 63: 100%|██████████| 449/449 [02:49<00:00,  2.65it/s, loss=0.00172, lr=9.91e-5, step=1795]\n",
      "\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "Epoch 64: 100%|██████████| 449/449 [02:45<00:00,  2.71it/s, loss=0.00325, lr=9.84e-5, step=2244]\n",
      "Epoch 65: 100%|██████████| 449/449 [02:44<00:00,  3.61it/s, loss=0.0225, lr=9.75e-5, step=2693] \n",
      "Epoch 65: 100%|██████████| 449/449 [02:44<00:00,  2.73it/s, loss=0.0225, lr=9.75e-5, step=2693]\n",
      "\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "Epoch 66: 100%|██████████| 449/449 [02:46<00:00,  2.69it/s, loss=0.143, lr=9.64e-5, step=3142]\n",
      "Epoch 67: 100%|██████████| 449/449 [02:48<00:00,  3.61it/s, loss=0.000732, lr=9.51e-5, step=3591]\n",
      "Epoch 67: 100%|██████████| 449/449 [02:48<00:00,  2.67it/s, loss=0.000732, lr=9.51e-5, step=3591]\n",
      "\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "Epoch 68: 100%|██████████| 449/449 [02:47<00:00,  2.69it/s, loss=0.0112, lr=9.36e-5, step=4040]\n",
      "Epoch 69: 100%|██████████| 449/449 [02:47<00:00,  3.61it/s, loss=0.000471, lr=9.2e-5, step=4489]\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "100%|██████████| 1000/1000 [01:48<00:00,  9.25it/s]\n",
      "\n",
      "Epoch 69: 100%|██████████| 449/449 [04:36<00:00,  1.62it/s, loss=0.000471, lr=9.2e-5, step=4489]\n",
      "\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A\n",
      "\u001b[A"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[22], line 3\u001b[0m\n\u001b[1;32m      1\u001b[0m args \u001b[38;5;241m=\u001b[39m (config, model, noise_scheduler, optimizer, train_dataloader, lr_scheduler, start_epoch)\n\u001b[0;32m----> 3\u001b[0m \u001b[43mnotebook_launcher\u001b[49m\u001b[43m(\u001b[49m\u001b[43mfine_tune_train_loop\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mnum_processes\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/accelerate/launchers.py:266\u001b[0m, in \u001b[0;36mnotebook_launcher\u001b[0;34m(function, args, num_processes, mixed_precision, use_port, master_addr, node_rank, num_nodes, rdzv_backend, rdzv_endpoint, rdzv_conf, rdzv_id, max_restarts, monitor_interval, log_line_prefix_template)\u001b[0m\n\u001b[1;32m    264\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m    265\u001b[0m     \u001b[38;5;28mprint\u001b[39m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mLaunching training on CPU.\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m--> 266\u001b[0m \u001b[43mfunction\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\n",
      "Cell \u001b[0;32mIn[17], line 54\u001b[0m, in \u001b[0;36mfine_tune_train_loop\u001b[0;34m(config, model, noise_scheduler, optimizer, train_dataloader, lr_scheduler, start_epoch)\u001b[0m\n\u001b[1;32m     52\u001b[0m noise_pred \u001b[38;5;241m=\u001b[39m model(noisy_images, timesteps, return_dict\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mFalse\u001b[39;00m)[\u001b[38;5;241m0\u001b[39m]\n\u001b[1;32m     53\u001b[0m loss \u001b[38;5;241m=\u001b[39m F\u001b[38;5;241m.\u001b[39mmse_loss(noise_pred, noise)\n\u001b[0;32m---> 54\u001b[0m \u001b[43maccelerator\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbackward\u001b[49m\u001b[43m(\u001b[49m\u001b[43mloss\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     56\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m accelerator\u001b[38;5;241m.\u001b[39msync_gradients:\n\u001b[1;32m     57\u001b[0m     accelerator\u001b[38;5;241m.\u001b[39mclip_grad_norm_(model\u001b[38;5;241m.\u001b[39mparameters(), \u001b[38;5;241m1.0\u001b[39m)\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/accelerate/accelerator.py:2155\u001b[0m, in \u001b[0;36mAccelerator.backward\u001b[0;34m(self, loss, **kwargs)\u001b[0m\n\u001b[1;32m   2153\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m\n\u001b[1;32m   2154\u001b[0m \u001b[38;5;28;01melif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mscaler \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[0;32m-> 2155\u001b[0m     \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mscaler\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mscale\u001b[49m\u001b[43m(\u001b[49m\u001b[43mloss\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbackward\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   2156\u001b[0m \u001b[38;5;28;01melif\u001b[39;00m learning_rate \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mhas_lomo_optimizer:\n\u001b[1;32m   2157\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mlomo_backward(loss, learning_rate)\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/_tensor.py:521\u001b[0m, in \u001b[0;36mTensor.backward\u001b[0;34m(self, gradient, retain_graph, create_graph, inputs)\u001b[0m\n\u001b[1;32m    511\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m has_torch_function_unary(\u001b[38;5;28mself\u001b[39m):\n\u001b[1;32m    512\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m handle_torch_function(\n\u001b[1;32m    513\u001b[0m         Tensor\u001b[38;5;241m.\u001b[39mbackward,\n\u001b[1;32m    514\u001b[0m         (\u001b[38;5;28mself\u001b[39m,),\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    519\u001b[0m         inputs\u001b[38;5;241m=\u001b[39minputs,\n\u001b[1;32m    520\u001b[0m     )\n\u001b[0;32m--> 521\u001b[0m \u001b[43mtorch\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mautograd\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbackward\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m    522\u001b[0m \u001b[43m    \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mgradient\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mretain_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcreate_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43minputs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43minputs\u001b[49m\n\u001b[1;32m    523\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/autograd/__init__.py:289\u001b[0m, in \u001b[0;36mbackward\u001b[0;34m(tensors, grad_tensors, retain_graph, create_graph, grad_variables, inputs)\u001b[0m\n\u001b[1;32m    284\u001b[0m     retain_graph \u001b[38;5;241m=\u001b[39m create_graph\n\u001b[1;32m    286\u001b[0m \u001b[38;5;66;03m# The reason we repeat the same comment below is that\u001b[39;00m\n\u001b[1;32m    287\u001b[0m \u001b[38;5;66;03m# some Python versions print out the first line of a multi-line function\u001b[39;00m\n\u001b[1;32m    288\u001b[0m \u001b[38;5;66;03m# calls in the traceback and some print out the last line\u001b[39;00m\n\u001b[0;32m--> 289\u001b[0m \u001b[43m_engine_run_backward\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m    290\u001b[0m \u001b[43m    \u001b[49m\u001b[43mtensors\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    291\u001b[0m \u001b[43m    \u001b[49m\u001b[43mgrad_tensors_\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    292\u001b[0m \u001b[43m    \u001b[49m\u001b[43mretain_graph\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    293\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcreate_graph\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    294\u001b[0m \u001b[43m    \u001b[49m\u001b[43minputs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    295\u001b[0m \u001b[43m    \u001b[49m\u001b[43mallow_unreachable\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m,\u001b[49m\n\u001b[1;32m    296\u001b[0m \u001b[43m    \u001b[49m\u001b[43maccumulate_grad\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m,\u001b[49m\n\u001b[1;32m    297\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/DIS2024/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:768\u001b[0m, in \u001b[0;36m_engine_run_backward\u001b[0;34m(t_outputs, *args, **kwargs)\u001b[0m\n\u001b[1;32m    766\u001b[0m     unregister_hooks \u001b[38;5;241m=\u001b[39m _register_logging_hooks_on_whole_graph(t_outputs)\n\u001b[1;32m    767\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[0;32m--> 768\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mVariable\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_execution_engine\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mrun_backward\u001b[49m\u001b[43m(\u001b[49m\u001b[43m  \u001b[49m\u001b[38;5;66;43;03m# Calls into the C++ engine to run the backward pass\u001b[39;49;00m\n\u001b[1;32m    769\u001b[0m \u001b[43m        \u001b[49m\u001b[43mt_outputs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\n\u001b[1;32m    770\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m  \u001b[38;5;66;03m# Calls into the C++ engine to run the backward pass\u001b[39;00m\n\u001b[1;32m    771\u001b[0m \u001b[38;5;28;01mfinally\u001b[39;00m:\n\u001b[1;32m    772\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m attach_logging_hooks:\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "args = (config, model, noise_scheduler, optimizer, train_dataloader, lr_scheduler, start_epoch)\n",
    "\n",
    "notebook_launcher(fine_tune_train_loop, args, num_processes=1)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "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": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}