self_attention_visualization.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "❗❗❗ Be sure using correct python environment, the python environment are different for methods \n",
      "🔄 Building Stable-Diffusion-Inpainting FreeCond control...\n",
      "# Load backbone model\n",
      "## Load backbone model = Stable Diffusion Inpainting 1.5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "vae/diffusion_pytorch_model.safetensors not found\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8ec63ceec656455686a9b2df65e04138",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Loading pipeline components...:   0%|          | 0/7 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "An error occurred while trying to fetch /nfs/home/bluedyee/.cache/huggingface/hub/models--sd-legacy--stable-diffusion-inpainting/snapshots/8a4288a76071f7280aedbdb3253bdb9e9d5d84bb/vae: Error no file named diffusion_pytorch_model.safetensors found in directory /nfs/home/bluedyee/.cache/huggingface/hub/models--sd-legacy--stable-diffusion-inpainting/snapshots/8a4288a76071f7280aedbdb3253bdb9e9d5d84bb/vae.\n",
      "Defaulting to unsafe serialization. Pass `allow_pickle=False` to raise an error instead.\n",
      "/nfs/home/bluedyee/miniconda3/envs/hdpainter/lib/python3.9/site-packages/torch/_utils.py:831: UserWarning: TypedStorage is deprecated. It will be removed in the future and UntypedStorage will be the only storage class. This should only matter to you if you are using storages directly.  To access UntypedStorage directly, use tensor.untyped_storage() instead of tensor.storage()\n",
      "  return self.fget.__get__(instance, owner)()\n",
      "An error occurred while trying to fetch /nfs/home/bluedyee/.cache/huggingface/hub/models--sd-legacy--stable-diffusion-inpainting/snapshots/8a4288a76071f7280aedbdb3253bdb9e9d5d84bb/unet: Error no file named diffusion_pytorch_model.safetensors found in directory /nfs/home/bluedyee/.cache/huggingface/hub/models--sd-legacy--stable-diffusion-inpainting/snapshots/8a4288a76071f7280aedbdb3253bdb9e9d5d84bb/unet.\n",
      "Defaulting to unsafe serialization. Pass `allow_pickle=False` to raise an error instead.\n",
      "/nfs/home/bluedyee/miniconda3/envs/hdpainter/lib/python3.9/site-packages/transformers/tokenization_utils_base.py:1617: FutureWarning: `clean_up_tokenization_spaces` was not set. It will be set to `True` by default. This behavior will be deprecated in transformers v4.45, and will be then set to `False` by default. For more details check this issue: https://github.com/huggingface/transformers/issues/31884\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "#❗❗❗❗❗\n",
    "# This Code require extremely high RAM, around 100GB+\n",
    "# To direcyly run this code, plase beware the memory requirement first\n",
    "# ❗❗❗❗❗\n",
    "\n",
    "from freecond_src.freecond_utils import get_pipeline_forward\n",
    "from freecond_src.freecond import fc_config\n",
    "from diffusers import DDIMScheduler\n",
    "from PIL import Image\n",
    "import torch\n",
    "torch.cuda.set_device(0)\n",
    "pipeline, forward = get_pipeline_forward(method=\"sd\",variant=\"sd15\", device=\"cuda\")\n",
    "\n",
    "# pipeline, forward = get_pipeline_forward(method=\"cn\")\n",
    "\n",
    "# pipeline, forward = get_pipeline_forward(method=\"hdp\",variant=\"sd15\",device=\"cuda\")\n",
    "\n",
    "# pipeline, forward = get_pipeline_forward(method=\"pp\",device=\"cuda\")\n",
    "\n",
    "# pipeline, forward = get_pipeline_forward(method=\"bn\",device=\"cuda\")\n",
    "\n",
    "# pipeline, forward = get_pipeline_forward(method=\"sd\",variant=\"sdxl\",device=\"cuda\")\n",
    "\n",
    "# pipeline.scheduler=DDIMScheduler.from_config(pipeline.scheduler.config) # This was used during the experiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "from PIL import ImageFont\n",
    "from PIL import ImageDraw \n",
    "import torch\n",
    "import torch.nn.functional as F\n",
    "\n",
    "def vis_cross_atn_dict(cross_atn_dict, token_pos=1, step=0, layer=31, indicator=\"c\"):\n",
    "    soft_sim=cross_atn_dict[indicator][step][layer]\n",
    "    # sum heads\n",
    "    soft_sim=torch.sum(soft_sim, 0, keepdim=False) \n",
    "    atn_map=soft_sim[:,token_pos]\n",
    "    n=int(atn_map.shape[0]**0.5)\n",
    "    atn_map=atn_map.reshape((n,n))*255\n",
    "    return Image.fromarray(atn_map.numpy().astype(\"uint8\"))\n",
    "\n",
    "def vis_self_1c(self_atn_tensor, mask):\n",
    "    # (4096, 4096)\n",
    "    N, M = self_atn_tensor.shape\n",
    "    H = int(N**0.5)\n",
    "    W = int(M**0.5)\n",
    "    resized_mask = F.interpolate(mask.unsqueeze(0).unsqueeze(0).cpu().to(torch.float), size=(H, W), mode='bilinear', align_corners=False)\n",
    "    out_tensor = self_atn_tensor.clone()\n",
    "    flat_mask = torch.flatten(resized_mask).to(bool)\n",
    "    out_tensor[:,flat_mask]*=1\n",
    "    out_tensor[:,~flat_mask]*= 0\n",
    "    out_tensor=torch.sum(out_tensor, dim=1)\n",
    "    return torch.reshape(out_tensor, (H,W))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Vis with prompt control"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Activating VIS Attention Control...\n",
      "# Visualize cross attention (1, 3, 5, 7... layers) = True\n",
      "# Visualize self attention (0, 2, 4, 6 ... layers)= True\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/nfs/home/bluedyee/miniconda3/envs/hdpainter/lib/python3.9/site-packages/torch/nn/modules/conv.py:456: UserWarning: Applied workaround for CuDNN issue, install nvrtc.so (Triggered internally at ../aten/src/ATen/native/cudnn/Conv_v8.cpp:80.)\n",
      "  return F.conv2d(input, weight, bias, self.stride,\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e715a8603c3640449cbefe52dfc07b53",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/50 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/nfs/home/bluedyee/github_repo/INPAINTING/FreeCond/freecond_src/freecond.py:75: UserWarning: ComplexHalf support is experimental and many operators don't support it yet. (Triggered internally at ../aten/src/ATen/EmptyTensor.cpp:31.)\n",
      "  x_freq = fft.fftn(x, dim=(-2, -1))\n"
     ]
    }
   ],
   "source": [
    "from PIL import Image\n",
    "from diffusers import DDIMScheduler\n",
    "from diffusers.utils import load_image, make_image_grid\n",
    "from freecond_src.vis_attn_control import VISAttentionControl, regiter_attention_editor_diffusers\n",
    "\n",
    "###\n",
    "nprompt=\"word, bad quality, bad anatomy, ugly, mutation, blurry, error\"\n",
    "SEED=777\n",
    "mask=Image.open(\"./demo_data/mask_1_2.png\").convert(\"L\").resize((512,512))\n",
    "init_image=Image.open(\"./demo_data/img_1_2.jpg\").convert(\"RGB\").resize((512,512))\n",
    "prompt=\"A golden retriever wearing astronaut gear, in cyberpunk style\"\n",
    "nprompt=\"word, bad quality, bad anatomy, ugly, mutation, blurry, error\"\n",
    "\n",
    "# store self-attention hidden layer to CPU\n",
    "freecond_editor = VISAttentionControl(start_step=0,end_step=50, start_layer=0, end_layer=16, vis_self=True, vis_cross=True)\n",
    "regiter_attention_editor_diffusers(pipeline, freecond_editor)\n",
    "\n",
    "# Default to SDI\n",
    "fc_control=fc_config(change_step=0,fg_1=1,fg_2=1,bg_1=0,bg_2=0,hq_1=1,hq_2=1,lq_1=1,lq_2=1,fq_th=32)\n",
    "# Setting FreeCond control for SDI15\n",
    "torch.manual_seed(777)\n",
    "default_out = forward(fc_control,init_image=init_image,mask_image=mask,prompt=prompt,negative_prompt=nprompt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/nfs/home/bluedyee/miniconda3/envs/hdpainter/lib/python3.9/site-packages/diffusers/configuration_utils.py:140: FutureWarning: Accessing config attribute `vae_latent_channels` directly via 'VaeImageProcessor' object attribute is deprecated. Please access 'vae_latent_channels' over 'VaeImageProcessor's config object instead, e.g. 'scheduler.config.vae_latent_channels'.\n",
      "  deprecate(\"direct config name access\", \"1.0.0\", deprecation_message, standard_warn=False)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "###\n",
    "condition_mask, c_masked_image_latents = pipeline.prepare_mask_condition(init_image,mask)\n",
    "EDITOR=freecond_editor\n",
    "mask=condition_mask[0][0]\n",
    "step=25\n",
    "layer=0 # ❗ the self-attention layer is evenly indexed (0, 2, 4, 6.... 32)\n",
    "indicator=\"c\"\n",
    "###\n",
    "step=25\n",
    "attn = EDITOR.self_atn_dict[indicator][step][layer][0]\n",
    "attn_out_1c = vis_self_1c(attn,mask).numpy()\n",
    "\n",
    "plt.imshow(attn_out_1c , cmap='coolwarm')\n",
    "#plt.colorbar()  \n",
    "plt.axis('off')  \n",
    "plt.tight_layout(pad=0)\n",
    "plt.savefig(\"./demo_out/self_attn.png\", bbox_inches='tight', pad_inches=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "step=25\n",
    "layers=[0,2,6,28,30] # ❗ the self-attention layer is evenly indexed (0, 2, 4, 6.... 30)\n",
    "attn = EDITOR.self_atn_dict[indicator][step][layer][0]\n",
    "attn_out_1c = vis_self_1c(attn,mask).numpy()\n",
    "\n",
    "plt.figure(figsize=(12, 6))\n",
    "for i, layer in enumerate(layers):\n",
    "    plt.subplot(1,len(layers),i+1)\n",
    "    attn = EDITOR.self_atn_dict[indicator][step][layer][0]\n",
    "    attn_out_1c = vis_self_1c(attn,mask).numpy()\n",
    "    plt.imshow(attn_out_1c , cmap='coolwarm')\n",
    "    #plt.colorbar()  \n",
    "    plt.axis('off')  \n",
    "    plt.tight_layout(pad=0)\n",
    "plt.savefig(\"./demo_out/self_attn_multi.png\", bbox_inches='tight', pad_inches=0)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "hdpainter",
   "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": "3.9.19"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}