ensemble_CIFAR.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"provenance":[],"machine_shape":"hm","authorship_tag":"ABX9TyPpoIuGaC6Tn8JweWRBwpqs"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"},"accelerator":"GPU","widgets":{"application/vnd.jupyter.widget-state+json":{"2a6e0b305a904b06bb7e46979925c145":{"model_module":"@jupyter-widgets/controls","model_name":"HBoxModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_42bb74f3e9b04e94ac96dac572cea174","IPY_MODEL_2a2101b5cb2340c88ec4d2785ed99892","IPY_MODEL_61b308013d2c481faa09d8d96dc0eeba"],"layout":"IPY_MODEL_4fd336a150f0401496687f98cb785b61"}},"42bb74f3e9b04e94ac96dac572cea174":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_6afd8820f4fc454290f26212d1e4001b","placeholder":"​","style":"IPY_MODEL_0e1be5d935a046989eb96305ac877729","value":"100%"}},"2a2101b5cb2340c88ec4d2785ed99892":{"model_module":"@jupyter-widgets/controls","model_name":"FloatProgressModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_cc59f138045a493db270f3837ffd8a21","max":170498071,"min":0,"orientation":"horizontal","style":"IPY_MODEL_7daa11ca50fd4f9e8a5f2a84f7f04861","value":170498071}},"61b308013d2c481faa09d8d96dc0eeba":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_7d399c8145944cdfbff855e0eb774825","placeholder":"​","style":"IPY_MODEL_e5c935af0e5c4e6a82828b4e9802ce9b","value":" 170498071/170498071 [00:13&lt;00:00, 14178929.65it/s]"}},"4fd336a150f0401496687f98cb785b61":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"6afd8820f4fc454290f26212d1e4001b":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"0e1be5d935a046989eb96305ac877729":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"cc59f138045a493db270f3837ffd8a21":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"7daa11ca50fd4f9e8a5f2a84f7f04861":{"model_module":"@jupyter-widgets/controls","model_name":"ProgressStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"7d399c8145944cdfbff855e0eb774825":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"e5c935af0e5c4e6a82828b4e9802ce9b":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}}}}},"cells":[{"cell_type":"code","source":["%pylab inline\n","%config InlineBackend.figure_format = 'retina'\n","%pip install torchmetrics\n"],"metadata":{"id":"7zW_pBWuHAPv","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1677135613175,"user_tz":480,"elapsed":6757,"user":{"displayName":"Bo Zhao","userId":"08721035053344232673"}},"outputId":"e8c6c237-ed49-4431-fd81-681463527af2"},"execution_count":1,"outputs":[{"output_type":"stream","name":"stdout","text":["Populating the interactive namespace from numpy and matplotlib\n","Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n","Collecting torchmetrics\n","  Downloading torchmetrics-0.11.1-py3-none-any.whl (517 kB)\n","\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m517.2/517.2 KB\u001b[0m \u001b[31m5.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n","\u001b[?25hRequirement already satisfied: typing-extensions in /usr/local/lib/python3.8/dist-packages (from torchmetrics) (4.5.0)\n","Requirement already satisfied: packaging in /usr/local/lib/python3.8/dist-packages (from torchmetrics) (23.0)\n","Requirement already satisfied: numpy>=1.17.2 in /usr/local/lib/python3.8/dist-packages (from torchmetrics) (1.22.4)\n","Requirement already satisfied: torch>=1.8.1 in /usr/local/lib/python3.8/dist-packages (from torchmetrics) (1.13.1+cu116)\n","Installing collected packages: torchmetrics\n","Successfully installed torchmetrics-0.11.1\n"]}]},{"cell_type":"code","source":["import torch\n","from torch import nn\n","import torch.nn.functional as F\n","import torchvision\n","from torchmetrics import Accuracy\n","import numpy as np\n","import matplotlib.pyplot as plt\n","import matplotlib.ticker as ticker\n","import os\n","import gc\n","\n","V = lambda x: x.detach().cpu().numpy()\n","\n","if not os.path.exists('figures'):\n","    os.mkdir('figures')\n","if not os.path.exists('figures/ensemble'):\n","    os.mkdir('figures/ensemble')\n","\n","torch.__version__"],"metadata":{"id":"ZYZc6Mm0HKh1","colab":{"base_uri":"https://localhost:8080/","height":35},"executionInfo":{"status":"ok","timestamp":1677135617738,"user_tz":480,"elapsed":4567,"user":{"displayName":"Bo Zhao","userId":"08721035053344232673"}},"outputId":"a714e40e-3e68-4bb2-b160-8a55bd012d75"},"execution_count":2,"outputs":[{"output_type":"execute_result","data":{"text/plain":["'1.13.1+cu116'"],"application/vnd.google.colaboratory.intrinsic+json":{"type":"string"}},"metadata":{},"execution_count":2}]},{"cell_type":"code","source":["class Trainer:\n","    def __init__(self, model, dataset, device = 'cpu'):\n","        # val_data = dataset.data[-n_val:]\n","        # val_targets = dataset.targets[-n_val:]\n","        self.device = device\n","        self.model = model.to(device)\n","        self.dataset = dataset\n","        self.dataset.data = self.dataset.data.to(device)\n","        self.dataset.targets = self.dataset.targets.to(device)\n","        self.optim = torch.optim.Adam(self.model.parameters())\n","        self.history = {'train_loss': [], 'train_acc':[], 'val_loss':[], 'val_acc':[]}\n","        self.accuracy = Accuracy(task='multiclass', num_classes=len(self.dataset.classes)).to(device)\n","        \n","    def fit(self, epochs =1, batch_size = 32, val_split=0.2, val_num = 1000, verbose=True):\n","        n = len(self.dataset.data)\n","        n_val = int(val_split*n)\n","        n_train = n-n_val \n","        \n","        for ep in range(epochs):\n","            total_loss = []\n","            acc_acc = []\n","            for i in range(0,n_train, batch_size):\n","                x,y = self.dataset.batch_get(i,i+batch_size)\n","\n","                self.optim.zero_grad()\n","                y_pred = self.model(x)\n","                loss = F.cross_entropy(y_pred, y)\n","                loss.backward()\n","                total_loss += [loss.item()+0]\n","                acc = self.accuracy(y_pred,y)\n","                acc_acc += [acc.item()+0]\n","                self.optim.step()\n","                if i % 20*batch_size and verbose:\n","                    print(f'{ep} Train: {i/n_train*100:.0f}\\t loss: {loss:.3g},\\t acc:{acc:.3f}', end='\\r')\n","            self.history['train_loss'] += [mean(total_loss)]\n","            self.history['train_acc'] += [mean(acc_acc)]\n","            x,y = self.dataset.batch_get(-min(n_val, val_num), n)\n","            y_pred = self.model(x)\n","            loss = F.cross_entropy(y_pred, y)\n","            acc = self.accuracy(y_pred,y)\n","            if verbose:\n","                print(f'\\n  Test:  \\t loss: {loss:.3g},\\t acc:{acc:.3f}')\n","            self.history['val_loss'] += [loss.item()+0]\n","            self.history['val_acc'] += [acc.item()+0]\n","        "],"metadata":{"id":"jdVnzmAmHML_","executionInfo":{"status":"ok","timestamp":1677135617739,"user_tz":480,"elapsed":7,"user":{"displayName":"Bo Zhao","userId":"08721035053344232673"}}},"execution_count":3,"outputs":[]},{"cell_type":"markdown","source":["# Group Action Activation layer\n","This layer implements the symmetry group action \n","$$ U\\sigma(H) \\to U\\pi(g,H)\\sigma(gH) $$\n","by wrapping it around an activation function $\\sigma' (x) = \\pi(g,x)\\sigma(gx) $, so that $U\\sigma'(H) =  U\\pi(g,H)\\sigma(gH)$. "],"metadata":{"id":"o_9HrYgfHNdN"}},{"cell_type":"code","source":["from pickle import FALSE\n","import warnings\n","\n","class GroupActivation(nn.Module):\n","    def __init__(self, in_features, nonlinearity = nn.Identity(), num_g = 1, eps = 1, T=None):\n","        \"\"\"Replaces a nonlinear activation layer. \n","        In normal mode, self(x) = nonlinearity(x)\n","        \n","        Args:\n","        in_features: # features of input; expects input of shape (batch, in_features, ...)\n","        num_g: number of group elements to use transform output\n","        eps: used in `g=exp(eps*T)` where `T` are gl_h Lie algebra elements. \n","        T: if None, random T (of Lie algebra) will be created. If provided, with shape (num_g,h,h) \n","            they will be used for computing g.   \n","        \n","        Flags:\n","        self.UPDATE_MODE: = False. If True, during forward pass, `num_g` batches of data\n","            will be stored as self.H for computing group rep $\\pi(g,H) = \\sigma(H)\\sigma(gH)^+$. \n","            self.H batch sizes are capped to in_features, as required for $pi(g,H)$.  \n","        self.G_ACT_MODE: = False. If True, uses group in forward pass \n","            return size : batch_size,... -> num_g*batch_size,... \n","        \"\"\"\n","        super().__init__()\n","        self.in_features = in_features\n","        self.nonlinearity = nonlinearity\n","        self.num_g = num_g \n","        self.eps = eps\n","        # in UPDATE_MODE, \n","        self.UPDATE_MODE = False\n","        self.G_ACT_MODE_RANDOM = False # if True, uses random transform in forward pass \n","        self.G_ACT_MODE_RANDOM_PERMUTE = False\n","        self.G_ACT_MODE_RANDOM_g = False\n","        self.PERMUTE_INTER = False\n","        self.invert_g_computed = False\n","        \n","\n","        self.G_ACT_MODE = False # if True, uses group in forward pass \n","        # in G-act mode: (batch_size -> num_g*batch_size)\n","        # self._G_INIT = False # set first time self.g is initialized \n","        self.init_G(T)\n","        self.H = None\n","        \n","    def group_act_lower(self,H):\n","        \"\"\" input H -> gH \"\"\"\n","        return torch.einsum('ncd...,bc...->nbd...', self.g, H)\n","    \n","    def batch_g_act(self,g,H):\n","        \"\"\"input and output shape (num_g, batch, chan_out, ...)\"\"\"\n","        # return torch.einsum('ndc...,nbc...->nbd...', g, H)\n","        return torch.einsum('ncd...,nbc...->nbd...', g, H)\n","    \n","    def group_act_upper(self,sigma_gH):\n","        \"\"\" sigma(gH) -> pi(g,H0)*sigma(gH) \n","        H0: batch used for computing pi(g,H0)\n","        H : new input \n","        input and output shape (num_g, batch, chan_out, ...)\n","        \"\"\"\n","        if self.G_ACT_MODE_RANDOM:\n","            I_diag = torch.ones(self.pi_g_H.size()[0], self.pi_g_H.size()[1])\n","            I = torch.diag_embed(I_diag)\n","            M = torch.rand_like(I)\n","            rand_M = I + self.eps * M\n","            rand_M = rand_M.to(DEVICE)\n","            return self.batch_g_act(rand_M, sigma_gH)\n","        elif self.G_ACT_MODE_RANDOM_PERMUTE:\n","            r = torch.randperm(self.pi_g_H.size()[0])\n","            c = torch.randperm(self.pi_g_H.size()[1])\n","            rand_pi_g_H = self.pi_g_H[r][:,c]\n","            r = torch.randperm(self.pi_g_H.size()[0])\n","            c = torch.randperm(self.pi_g_H.size()[1])\n","            rand_pi_g_H = rand_pi_g_H[r][:,c]\n","            rand_pi_g_H = rand_pi_g_H.to(DEVICE)\n","            return self.batch_g_act(rand_pi_g_H, sigma_gH)\n","        elif self.G_ACT_MODE_RANDOM_g:\n","            if not self.invert_g_computed:\n","                self.g_inv = torch.linalg.pinv(self.g)\n","                self.invert_g_computed = True\n","            return self.batch_g_act(self.g_inv, sigma_gH)\n","        else:\n","            return self.batch_g_act(self.pi_g_H, sigma_gH)\n","    \n","    \n","    def get_group_rep_upper(self,):\n","        \"\"\"Compute $\\pi(g,H) = \\sigma(H) \\sigma(gH)^+$ \n","        Because the batch index is first (instead of last in our theory)\n","        we actually need `pi(g,H) = pinv(sigma(gH)) @ sigma(H)`. \n","        We use torch.linalg.lstsq(A,B) = A.pinverse() @ B \n","        (supposedly more numerically stable than pinverse). \n","        \"\"\"\n","        # to use lstsq, batch, channel must be last two dims\n","        # to (num_g, ..., batch, cin)  \n","        H_perm = self.H.permute(0,*torch.arange(3,self.H.ndim),1,2)\n","        H_perm = self.nonlinearity(H_perm)\n","        \n","        # stored H are (num_g, batch, cin, ...); regular H is (batch, cin,...)\n","        gH = self.batch_g_act(self.g, self.H)\n","        # to (num_g, ..., batch, channel)  \n","        gH_perm = gH.permute(0,*torch.arange(3,gH.ndim),1,2)\n","        gH_perm = self.nonlinearity(gH_perm)\n","        \n","        self.pi_g_H = torch.linalg.lstsq(gH_perm, H_perm).solution\n","        # move the (in,out) indices of g first\n","        # shape (num_g, cin, cin, ...)\n","        self.pi_g_H = self.pi_g_H.permute(0,-2,-1,*torch.arange(1, self.pi_g_H.ndim-2))\n","                \n","    def store_batch(self,H, noise=1e-4):\n","        \"\"\"stores a queue of H of size num_g. \n","        Note: because we need in_features >= batch, the batch size is capped to in_features.\n","        \"\"\"\n","        H_capped = H[:self.in_features]\n","        \n","        if not hasattr(self,'_H_counter'):\n","            # initialize \n","            self._H_counter = 0\n","            self.H = noise*torch.randn(self.num_g, *H_capped.shape).to(H_capped.device)\n","        \n","        self.H[self._H_counter] += H_capped\n","        # print(f'H batch {self._H_counter}/{self.num_g} stored')\n","        self._H_counter = (self._H_counter + 1) % self.num_g \n","        \n","    def update_g(self,H):\n","        self.store_batch(H)\n","        if self._H_counter == 0:\n","            warnings.warn(\"Suffiecient H stored. Computing pi(g,H). \"\n","                          \"Consider setting self.UPDATE_MODE = False\")\n","            self.get_group_rep_upper() \n","        \n","    def forward(self, H):\n","        if self.UPDATE_MODE:\n","            # store batches \n","            self.update_g(H)\n","            \n","        if not self.G_ACT_MODE: \n","            return self.nonlinearity(H)\n","        elif self.PERMUTE_INTER:\n","            # [0.5(I+g)]^{-1} \\sigma [0.5(I+g) H]\n","            # g = (I+ eps*0.5*(I+perm))/(1+eps)\n","            n = self.g.size()[0]\n","            h = H.size()[1]\n","            g = torch.zeros_like(self.g)\n","            g_inv = torch.zeros_like(self.g)\n","\n","            for i in range(n):\n","                perm = torch.argsort(torch.rand(h))\n","                perm_inv = torch.argsort(perm)\n","                I = torch.eye(h)\n","                g[i] = (I + 0.5 * self.eps * (I + I[perm])) / (1 + self.eps)\n","                g_inv[i] = torch.linalg.pinv(g[i])\n","            gH = torch.einsum('ncd...,bc...->nbd...', g, H)\n","            gH = self.nonlinearity(gH)\n","            gH = torch.einsum('ncd...,nbc...->nbd...', g_inv, gH)\n","            return gH.flatten(start_dim=0, end_dim=1)\n","        else:\n","            gH = self.group_act_lower(H)\n","            gH = self.nonlinearity(gH)\n","            gH = self.group_act_upper(gH)\n","            # combine num_g and batch \n","            return gH.flatten(start_dim=0, end_dim=1)\n","    \n","    def transform_U(self,):\n","        \"\"\" upper weight U -> g U = U pi(g,H) \"\"\"\n","        pass \n","    \n","    def init_G(self,T=None):\n","        self.T = T\n","        cin = self.in_features #self.H.shape[2] #(num_g, batch, cin, ...)\n","        \n","        if type(self.T) == type(None):\n","            if self.num_g > 0:\n","                self.T = torch.randn(self.num_g, cin, cin) / cin**.5\n","            else:\n","                self.T = torch.zeros(1,  cin, cin)\n","            \n","        else:\n","            assert type(self.T)==torch.Tensor\n","            self.num_g = len(self.T)\n","            \n","        # self.g = torch.eye(cin) + self.eps * self.T   \n","        self.g = torch.linalg.matrix_exp(self.eps * self.T)\n","        # self.g = torch.rand(self.num_g,cin,cin)\n","        self.T = nn.Parameter(self.T)\n","        self.g = nn.Parameter(self.g)\n","        \n","        self.T.requires_grad = False\n","        self.g.requires_grad = False\n","        "],"metadata":{"id":"ZgUqQPMPHPBi","executionInfo":{"status":"ok","timestamp":1677135617739,"user_tz":480,"elapsed":6,"user":{"displayName":"Bo Zhao","userId":"08721035053344232673"}}},"execution_count":4,"outputs":[]},{"cell_type":"markdown","source":["# cifar10"],"metadata":{"id":"zfZ1JsyezB29"}},{"cell_type":"code","source":["DEVICE = 'cuda'"],"metadata":{"id":"yweg0TTDy2Ry","executionInfo":{"status":"ok","timestamp":1677135618142,"user_tz":480,"elapsed":409,"user":{"displayName":"Bo Zhao","userId":"08721035053344232673"}}},"execution_count":5,"outputs":[]},{"cell_type":"code","source":["# cifar10 = torchvision.datasets.CIFAR10('../data/',download=False)\n","# the data from CIFAR10 class is in PIL format and needs processing before we can use it in a pytorch model\n","\n","class CIFAR10(torchvision.datasets.CIFAR10):\n","    def __init__(self, device='cpu', **kws):\n","        super().__init__(**kws)\n","        self.data = torch.tensor(self.data.transpose(0,3,1,2)/255. - 0.5, dtype=torch.float32).to(device)\n","        self.targets = torch.tensor(self.targets, dtype = torch.long).to(device)\n","\n","    def __getitem__(self,i):\n","        return (self.data[i], self.targets[i])\n","    \n","    def batch_get(self,i,j):\n","        return (self.data[i:j], self.targets[i:j])\n","\n","    \n","cif2= CIFAR10(root='../data/', download=True, device=DEVICE)\n","cif2_test = CIFAR10(root='../data/', train=False, device=DEVICE)\n","len(cif2_test.data), len(cif2.data)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":103,"referenced_widgets":["2a6e0b305a904b06bb7e46979925c145","42bb74f3e9b04e94ac96dac572cea174","2a2101b5cb2340c88ec4d2785ed99892","61b308013d2c481faa09d8d96dc0eeba","4fd336a150f0401496687f98cb785b61","6afd8820f4fc454290f26212d1e4001b","0e1be5d935a046989eb96305ac877729","cc59f138045a493db270f3837ffd8a21","7daa11ca50fd4f9e8a5f2a84f7f04861","7d399c8145944cdfbff855e0eb774825","e5c935af0e5c4e6a82828b4e9802ce9b"]},"id":"-kCxT5M7zDhG","executionInfo":{"status":"ok","timestamp":1677135643253,"user_tz":480,"elapsed":25115,"user":{"displayName":"Bo Zhao","userId":"08721035053344232673"}},"outputId":"b3ac0915-6c79-45df-f00a-0a152bb7c5a7"},"execution_count":6,"outputs":[{"output_type":"stream","name":"stdout","text":["Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to ../data/cifar-10-python.tar.gz\n"]},{"output_type":"display_data","data":{"text/plain":["  0%|          | 0/170498071 [00:00<?, ?it/s]"],"application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"2a6e0b305a904b06bb7e46979925c145"}},"metadata":{}},{"output_type":"stream","name":"stdout","text":["Extracting ../data/cifar-10-python.tar.gz to ../data/\n"]},{"output_type":"execute_result","data":{"text/plain":["(10000, 50000)"]},"metadata":{},"execution_count":6}]},{"cell_type":"code","source":["m = nn.LeakyReLU(0.1)\n","def relu_sqr(x):\n","    return m(x) * m(x) + 1e-2 * x"],"metadata":{"id":"wwgvEbGx0Hu6","executionInfo":{"status":"ok","timestamp":1677135643253,"user_tz":480,"elapsed":24,"user":{"displayName":"Bo Zhao","userId":"08721035053344232673"}}},"execution_count":7,"outputs":[]},{"cell_type":"code","source":["h_list = [32] #[4, 8, 16, 32, 64]\n","num_g_list = [10, 20, 40, 60, 80, 100, 120]\n","eps_list = [2e1, 1e1, 1e0, 1e-1, 1e-2, 1e-3, 1e-4]\n","\n","acc_original = np.zeros((len(h_list), len(num_g_list), len(eps_list)))\n","acc_group = np.zeros((len(h_list), len(num_g_list), len(eps_list)))\n","acc_random = np.zeros((len(h_list), len(num_g_list), len(eps_list)))\n","acc_random_permute = np.zeros((len(h_list), len(num_g_list), len(eps_list)))\n","acc_random_g = np.zeros((len(h_list), len(num_g_list), len(eps_list)))\n","acc_random_inter = np.zeros((len(h_list), len(num_g_list), len(eps_list)))\n","val_loss_original = np.zeros((len(h_list), len(num_g_list), len(eps_list)))\n","val_loss_group = np.zeros((len(h_list), len(num_g_list), len(eps_list)))\n","val_loss_random = np.zeros((len(h_list), len(num_g_list), len(eps_list)))\n","val_loss_random_permute = np.zeros((len(h_list), len(num_g_list), len(eps_list)))\n","val_loss_random_g = np.zeros((len(h_list), len(num_g_list), len(eps_list)))\n","val_loss_random_inter = np.zeros((len(h_list), len(num_g_list), len(eps_list)))"],"metadata":{"id":"H3AxDZp3_Kmd","executionInfo":{"status":"ok","timestamp":1677135643253,"user_tz":480,"elapsed":22,"user":{"displayName":"Bo Zhao","userId":"08721035053344232673"}}},"execution_count":8,"outputs":[]},{"cell_type":"code","source":["#im-size \n","p = 32\n","cin = 3\n","nf = 32\n","\n","h_list = [16] #[4, 8, 16, 32, 64]\n","num_g_list = [10, 20, 40, 60, 80, 100, 120]\n","eps_list = [2e1, 1e1, 1e0, 1e-1, 1e-2, 1e-3, 1e-4]\n","\n","for h_idx, h in enumerate(h_list):\n","    for num_g_idx, num_g in enumerate(num_g_list): #[(6, 120)]: #, (6, 120) (2, 1e0), \n","        for eps_idx, eps in enumerate(eps_list): #[(0, 2e1), (1, 1e1), (2, 1e0), (3, 1e-1), (4, 1e-3), (5, 1e-3), (6, 1e-4)]:\n","            ga = GroupActivation(h, num_g=num_g, nonlinearity=relu_sqr, eps=eps)\n","\n","            model = nn.Sequential(\n","                nn.Conv2d(cin,nf,3, padding='same'),\n","                nn.LeakyReLU(),\n","                nn.MaxPool2d(2,2),\n","                nn.Flatten(),\n","                nn.Linear((p//2)**2*nf, h),\n","                ga,\n","                nn.Linear(h,10),\n","                nn.Softmax(1),\n","            ).to(DEVICE)\n","            model.trainer = Trainer(model, cif2, device = DEVICE)\n","            model.trainer.fit(10, verbose=False)\n","            \n","            # group\n","            ga.UPDATE_MODE = True\n","            ga.G_ACT_MODE = False\n","\n","            x,y = cif2.batch_get(-ga.in_features*ga.num_g, len(cif2.data))\n","            h = ga.in_features\n","            for i in range(ga.num_g):\n","                yp = model(x[i*h:(i+1)*h])\n","                \n","            last = -ga.in_features*ga.num_g\n","            num = 1000\n","            x,y = cif2.batch_get(-num+last,last)\n","\n","            ga.UPDATE_MODE = False\n","            ga.G_ACT_MODE = False\n","            ga.G_ACT_MODE_RANDOM = False\n","            y_pred = model(x)\n","\n","            ga.G_ACT_MODE = True\n","            y_group = model(x)\n","\n","            ga.G_ACT_MODE_RANDOM = True\n","            ga.G_ACT_MODE_RANDOM_PERMUTE = False\n","            ga.G_ACT_MODE_RANDOM_g = False\n","            y_random = model(x)\n","\n","            ga.G_ACT_MODE_RANDOM = False\n","            ga.G_ACT_MODE_RANDOM_PERMUTE = True\n","            ga.G_ACT_MODE_RANDOM_g = False\n","            y_random_permute = model(x)\n","\n","            ga.G_ACT_MODE_RANDOM = False\n","            ga.G_ACT_MODE_RANDOM_PERMUTE = False\n","            ga.G_ACT_MODE_RANDOM_g = True\n","            y_random_g = model(x)\n","\n","            ga.G_ACT_MODE_RANDOM_g = False\n","            ga.PERMUTE_INTER = True\n","            y_random_inter = model(x)\n","\n","            accuracy = model.trainer.accuracy\n","            y_group = y_group.reshape(ga.num_g, *y_pred.shape )\n","            y_random = y_random.reshape(ga.num_g, *y_pred.shape )\n","            y_random_permute = y_random_permute.reshape(ga.num_g, *y_pred.shape )\n","            y_random_g = y_random_g.reshape(ga.num_g, *y_pred.shape )\n","            y_random_inter = y_random_inter.reshape(ga.num_g, *y_pred.shape )\n","\n","            val_loss_original[h_idx][num_g_idx][eps_idx] = F.cross_entropy(y_pred, y).item()\n","            val_loss_group[h_idx][num_g_idx][eps_idx] = F.cross_entropy(y_group.mean(0), y).item()\n","            val_loss_random[h_idx][num_g_idx][eps_idx] = F.cross_entropy(y_random.mean(0), y).item()\n","            val_loss_random_permute[h_idx][num_g_idx][eps_idx] = F.cross_entropy(y_random_permute.mean(0), y).item()\n","            val_loss_random_g[h_idx][num_g_idx][eps_idx] = F.cross_entropy(y_random_g.mean(0), y).item()\n","            val_loss_random_inter[h_idx][num_g_idx][eps_idx] = F.cross_entropy(y_random_inter.mean(0), y).item()\n","\n","            acc_original[h_idx][num_g_idx][eps_idx] = accuracy(y_pred, y)\n","            acc_group[h_idx][num_g_idx][eps_idx] = torch.mode(y_group.data.max(2)[1], 0).values.eq(y).long().cpu().sum() / y.size()[0]  # accuracy(y_group.mean(0), y)\n","            acc_random[h_idx][num_g_idx][eps_idx] = torch.mode(y_random.data.max(2)[1], 0).values.eq(y).long().cpu().sum() / y.size()[0]\n","            acc_random_permute[h_idx][num_g_idx][eps_idx] = torch.mode(y_random_permute.data.max(2)[1], 0).values.eq(y).long().cpu().sum() / y.size()[0]\n","            acc_random_g[h_idx][num_g_idx][eps_idx] = torch.mode(y_random_g.data.max(2)[1], 0).values.eq(y).long().cpu().sum() / y.size()[0]\n","            acc_random_inter[h_idx][num_g_idx][eps_idx] = torch.mode(y_random_inter.data.max(2)[1], 0).values.eq(y).long().cpu().sum() / y.size()[0]\n","\n","\n","            print(f'h={h} num_g={num_g} eps={eps} \\\n","                Accuracy: \\toriginal={accuracy(y_pred, y):.3f} \\\n","                \\tgroup:{accuracy(y_group.mean(0), y):.3f} \\\n","                \\trandom:{accuracy(y_random.mean(0), y):.3f} \\\n","                \\trandom_permute:{accuracy(y_random_permute.mean(0), y):.3f} \\\n","                \\trandom_g:{accuracy(y_random_g.mean(0), y):.3f} \\\n","                \\trandom_inter_g:{accuracy(y_random_inter.mean(0), y):.3f} \\\n","                Validation loss: \\toriginal={F.cross_entropy(y_pred, y).item():.6f} \\\n","                \\tgroup:{F.cross_entropy(y_group.mean(0), y).item():.6f} \\\n","                \\trandom:{F.cross_entropy(y_random.mean(0), y).item():.6f} \\\n","                \\trandom_permute:{F.cross_entropy(y_random_permute.mean(0), y).item():.6f} \\\n","                \\trandom_g:{F.cross_entropy(y_random_g.mean(0), y).item():.6f} \\\n","                \\trandom_inter_g:{F.cross_entropy(y_random_inter.mean(0), y).item():.6f}')\n","\n"],"metadata":{"id":"t7ABuSnJzkpp","executionInfo":{"status":"ok","timestamp":1677137587541,"user_tz":480,"elapsed":1944310,"user":{"displayName":"Bo Zhao","userId":"08721035053344232673"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"e43f8e2c-c53e-4e23-bed7-f9cf30273fc5"},"execution_count":9,"outputs":[{"output_type":"stream","name":"stderr","text":["<ipython-input-4-9a0d54ca6e1f>:124: UserWarning: Suffiecient H stored. Computing pi(g,H). Consider setting self.UPDATE_MODE = False\n","  warnings.warn(\"Suffiecient H stored. Computing pi(g,H). \"\n"]},{"output_type":"stream","name":"stdout","text":["h=16 num_g=10 eps=20.0                 Accuracy: \toriginal=0.560                 \tgroup:0.189                 \trandom:0.099                 \trandom_permute:0.094                 \trandom_g:0.086                 \trandom_inter_g:0.228                 Validation loss: \toriginal=1.894813                 \tgroup:2.270056                 \trandom:2.321812                 \trandom_permute:2.301889                 \trandom_g:2.317745                 \trandom_inter_g:2.251303\n","h=16 num_g=10 eps=10.0                 Accuracy: \toriginal=0.568                 \tgroup:0.219                 \trandom:0.111                 \trandom_permute:0.113                 \trandom_g:0.069                 \trandom_inter_g:0.283                 Validation loss: \toriginal=1.889618                 \tgroup:2.257545                 \trandom:2.309163                 \trandom_permute:2.311495                 \trandom_g:2.307151                 \trandom_inter_g:2.221727\n","h=16 num_g=10 eps=1.0                 Accuracy: \toriginal=0.539                 \tgroup:0.249                 \trandom:0.126                 \trandom_permute:0.103                 \trandom_g:0.088                 \trandom_inter_g:0.392                 Validation loss: \toriginal=1.914968                 \tgroup:2.224354                 \trandom:2.301146                 \trandom_permute:2.307095                 \trandom_g:2.304057                 \trandom_inter_g:2.077838\n","h=16 num_g=10 eps=0.1                 Accuracy: \toriginal=0.520                 \tgroup:0.503                 \trandom:0.447                 \trandom_permute:0.121                 \trandom_g:0.507                 \trandom_inter_g:0.510                 Validation loss: \toriginal=1.936608                 \tgroup:2.064263                 \trandom:2.040642                 \trandom_permute:2.305132                 \trandom_g:1.980121                 \trandom_inter_g:1.947233\n","h=16 num_g=10 eps=0.01                 Accuracy: \toriginal=0.499                 \tgroup:0.503                 \trandom:0.491                 \trandom_permute:0.025                 \trandom_g:0.505                 \trandom_inter_g:0.498                 Validation loss: \toriginal=1.958237                 \tgroup:1.974958                 \trandom:1.964005                 \trandom_permute:2.364100                 \trandom_g:1.959044                 \trandom_inter_g:1.958541\n","h=16 num_g=10 eps=0.001                 Accuracy: \toriginal=0.568                 \tgroup:0.565                 \trandom:0.567                 \trandom_permute:0.033                 \trandom_g:0.568                 \trandom_inter_g:0.568                 Validation loss: \toriginal=1.894590                 \tgroup:1.893947                 \trandom:1.894581                 \trandom_permute:2.411264                 \trandom_g:1.894649                 \trandom_inter_g:1.894583\n","h=16 num_g=10 eps=0.0001                 Accuracy: \toriginal=0.576                 \tgroup:0.575                 \trandom:0.576                 \trandom_permute:0.134                 \trandom_g:0.576                 \trandom_inter_g:0.576                 Validation loss: \toriginal=1.886407                 \tgroup:1.887736                 \trandom:1.886424                 \trandom_permute:2.318406                 \trandom_g:1.886417                 \trandom_inter_g:1.886403\n","h=16 num_g=20 eps=20.0                 Accuracy: \toriginal=0.473                 \tgroup:0.099                 \trandom:0.102                 \trandom_permute:0.099                 \trandom_g:0.108                 \trandom_inter_g:0.228                 Validation loss: \toriginal=1.987835                 \tgroup:nan                 \trandom:2.326374                 \trandom_permute:nan                 \trandom_g:2.306916                 \trandom_inter_g:2.235676\n","h=16 num_g=20 eps=10.0                 Accuracy: \toriginal=0.586                 \tgroup:0.216                 \trandom:0.062                 \trandom_permute:0.077                 \trandom_g:0.112                 \trandom_inter_g:0.376                 Validation loss: \toriginal=1.873331                 \tgroup:2.265661                 \trandom:2.323563                 \trandom_permute:2.301918                 \trandom_g:2.301995                 \trandom_inter_g:2.168680\n","h=16 num_g=20 eps=1.0                 Accuracy: \toriginal=0.551                 \tgroup:0.304                 \trandom:0.099                 \trandom_permute:0.076                 \trandom_g:0.151                 \trandom_inter_g:0.513                 Validation loss: \toriginal=1.903636                 \tgroup:2.223112                 \trandom:2.327670                 \trandom_permute:2.305523                 \trandom_g:2.283354                 \trandom_inter_g:2.077939\n","h=16 num_g=20 eps=0.1                 Accuracy: \toriginal=0.519                 \tgroup:0.524                 \trandom:0.460                 \trandom_permute:0.181                 \trandom_g:0.509                 \trandom_inter_g:0.514                 Validation loss: \toriginal=1.939404                 \tgroup:2.075357                 \trandom:2.020762                 \trandom_permute:2.291297                 \trandom_g:1.995878                 \trandom_inter_g:1.948139\n","h=16 num_g=20 eps=0.01                 Accuracy: \toriginal=0.456                 \tgroup:0.456                 \trandom:0.455                 \trandom_permute:0.105                 \trandom_g:0.455                 \trandom_inter_g:0.456                 Validation loss: \toriginal=2.002489                 \tgroup:2.007062                 \trandom:2.001978                 \trandom_permute:2.317892                 \trandom_g:2.001653                 \trandom_inter_g:2.002278\n","h=16 num_g=20 eps=0.001                 Accuracy: \toriginal=0.583                 \tgroup:0.582                 \trandom:0.582                 \trandom_permute:0.105                 \trandom_g:0.582                 \trandom_inter_g:0.583                 Validation loss: \toriginal=1.878447                 \tgroup:1.878216                 \trandom:1.878687                 \trandom_permute:2.348226                 \trandom_g:1.878506                 \trandom_inter_g:1.878401\n","h=16 num_g=20 eps=0.0001                 Accuracy: \toriginal=0.486                 \tgroup:0.486                 \trandom:0.486                 \trandom_permute:0.103                 \trandom_g:0.486                 \trandom_inter_g:0.486                 Validation loss: \toriginal=1.971995                 \tgroup:1.972056                 \trandom:1.971912                 \trandom_permute:2.354554                 \trandom_g:1.971969                 \trandom_inter_g:1.972016\n","h=16 num_g=40 eps=20.0                 Accuracy: \toriginal=0.527                 \tgroup:0.097                 \trandom:0.097                 \trandom_permute:0.097                 \trandom_g:0.135                 \trandom_inter_g:0.432                 Validation loss: \toriginal=1.930147                 \tgroup:nan                 \trandom:2.328264                 \trandom_permute:nan                 \trandom_g:2.302235                 \trandom_inter_g:2.121158\n","h=16 num_g=40 eps=10.0                 Accuracy: \toriginal=0.582                 \tgroup:0.295                 \trandom:0.096                 \trandom_permute:0.055                 \trandom_g:0.098                 \trandom_inter_g:0.303                 Validation loss: \toriginal=1.876277                 \tgroup:2.246070                 \trandom:2.324022                 \trandom_permute:2.305511                 \trandom_g:2.304455                 \trandom_inter_g:2.187646\n","h=16 num_g=40 eps=1.0                 Accuracy: \toriginal=0.580                 \tgroup:0.355                 \trandom:0.106                 \trandom_permute:0.106                 \trandom_g:0.078                 \trandom_inter_g:0.524                 Validation loss: \toriginal=1.879126                 \tgroup:2.217615                 \trandom:2.327694                 \trandom_permute:2.304461                 \trandom_g:2.300773                 \trandom_inter_g:2.033031\n","h=16 num_g=40 eps=0.1                 Accuracy: \toriginal=0.551                 \tgroup:0.547                 \trandom:0.477                 \trandom_permute:0.041                 \trandom_g:0.545                 \trandom_inter_g:0.550                 Validation loss: \toriginal=1.913052                 \tgroup:2.069220                 \trandom:2.009284                 \trandom_permute:2.306823                 \trandom_g:1.963248                 \trandom_inter_g:1.918817\n","h=16 num_g=40 eps=0.01                 Accuracy: \toriginal=0.491                 \tgroup:0.488                 \trandom:0.489                 \trandom_permute:0.107                 \trandom_g:0.489                 \trandom_inter_g:0.490                 Validation loss: \toriginal=1.966257                 \tgroup:1.983633                 \trandom:1.966667                 \trandom_permute:2.314230                 \trandom_g:1.966426                 \trandom_inter_g:1.965285\n","h=16 num_g=40 eps=0.001                 Accuracy: \toriginal=0.494                 \tgroup:0.494                 \trandom:0.495                 \trandom_permute:0.077                 \trandom_g:0.494                 \trandom_inter_g:0.494                 Validation loss: \toriginal=1.966254                 \tgroup:1.969748                 \trandom:1.965719                 \trandom_permute:2.369199                 \trandom_g:1.966149                 \trandom_inter_g:1.966327\n","h=16 num_g=40 eps=0.0001                 Accuracy: \toriginal=0.558                 \tgroup:0.558                 \trandom:0.558                 \trandom_permute:0.083                 \trandom_g:0.558                 \trandom_inter_g:0.558                 Validation loss: \toriginal=1.900743                 \tgroup:1.900754                 \trandom:1.900739                 \trandom_permute:2.374061                 \trandom_g:1.900735                 \trandom_inter_g:1.900745\n","h=16 num_g=60 eps=20.0                 Accuracy: \toriginal=0.520                 \tgroup:0.107                 \trandom:0.100                 \trandom_permute:0.107                 \trandom_g:0.075                 \trandom_inter_g:0.452                 Validation loss: \toriginal=1.936656                 \tgroup:nan                 \trandom:2.339796                 \trandom_permute:nan                 \trandom_g:2.304867                 \trandom_inter_g:2.215611\n","h=16 num_g=60 eps=10.0                 Accuracy: \toriginal=0.594                 \tgroup:0.377                 \trandom:0.103                 \trandom_permute:0.104                 \trandom_g:0.082                 \trandom_inter_g:0.391                 Validation loss: \toriginal=1.868580                 \tgroup:2.250380                 \trandom:2.318843                 \trandom_permute:2.304039                 \trandom_g:2.305661                 \trandom_inter_g:2.110497\n","h=16 num_g=60 eps=1.0                 Accuracy: \toriginal=0.529                 \tgroup:0.345                 \trandom:0.107                 \trandom_permute:0.103                 \trandom_g:0.073                 \trandom_inter_g:0.405                 Validation loss: \toriginal=1.935368                 \tgroup:2.232922                 \trandom:2.324842                 \trandom_permute:2.302105                 \trandom_g:2.305020                 \trandom_inter_g:2.044950\n","h=16 num_g=60 eps=0.1                 Accuracy: \toriginal=0.521                 \tgroup:0.518                 \trandom:0.472                 \trandom_permute:0.150                 \trandom_g:0.527                 \trandom_inter_g:0.524                 Validation loss: \toriginal=1.936486                 \tgroup:2.076239                 \trandom:2.020322                 \trandom_permute:2.293226                 \trandom_g:1.975987                 \trandom_inter_g:1.944088\n","h=16 num_g=60 eps=0.01                 Accuracy: \toriginal=0.461                 \tgroup:0.460                 \trandom:0.453                 \trandom_permute:0.170                 \trandom_g:0.462                 \trandom_inter_g:0.460                 Validation loss: \toriginal=1.999931                 \tgroup:2.017751                 \trandom:2.000295                 \trandom_permute:2.277372                 \trandom_g:1.998794                 \trandom_inter_g:1.999183\n","h=16 num_g=60 eps=0.001                 Accuracy: \toriginal=0.521                 \tgroup:0.521                 \trandom:0.521                 \trandom_permute:0.103                 \trandom_g:0.521                 \trandom_inter_g:0.521                 Validation loss: \toriginal=1.937319                 \tgroup:1.937632                 \trandom:1.937302                 \trandom_permute:2.342951                 \trandom_g:1.937291                 \trandom_inter_g:1.937392\n","h=16 num_g=60 eps=0.0001                 Accuracy: \toriginal=0.536                 \tgroup:0.536                 \trandom:0.537                 \trandom_permute:0.039                 \trandom_g:0.536                 \trandom_inter_g:0.536                 Validation loss: \toriginal=1.921896                 \tgroup:1.921903                 \trandom:1.921882                 \trandom_permute:2.416296                 \trandom_g:1.921897                 \trandom_inter_g:1.921895\n","h=16 num_g=80 eps=20.0                 Accuracy: \toriginal=0.521                 \tgroup:0.108                 \trandom:0.109                 \trandom_permute:0.108                 \trandom_g:0.098                 \trandom_inter_g:0.484                 Validation loss: \toriginal=1.933647                 \tgroup:nan                 \trandom:2.305378                 \trandom_permute:nan                 \trandom_g:2.304259                 \trandom_inter_g:2.036885\n","h=16 num_g=80 eps=10.0                 Accuracy: \toriginal=0.538                 \tgroup:0.205                 \trandom:0.108                 \trandom_permute:0.116                 \trandom_g:0.111                 \trandom_inter_g:0.369                 Validation loss: \toriginal=1.930063                 \tgroup:2.259768                 \trandom:2.312854                 \trandom_permute:2.304083                 \trandom_g:2.304441                 \trandom_inter_g:2.115943\n","h=16 num_g=80 eps=1.0                 Accuracy: \toriginal=0.527                 \tgroup:0.365                 \trandom:0.099                 \trandom_permute:0.076                 \trandom_g:0.097                 \trandom_inter_g:0.366                 Validation loss: \toriginal=1.931801                 \tgroup:2.234208                 \trandom:2.315311                 \trandom_permute:2.308906                 \trandom_g:2.307897                 \trandom_inter_g:2.074027\n","h=16 num_g=80 eps=0.1                 Accuracy: \toriginal=0.484                 \tgroup:0.485                 \trandom:0.468                 \trandom_permute:0.127                 \trandom_g:0.477                 \trandom_inter_g:0.485                 Validation loss: \toriginal=1.975952                 \tgroup:2.061550                 \trandom:2.004950                 \trandom_permute:2.319473                 \trandom_g:1.996605                 \trandom_inter_g:1.975209\n","h=16 num_g=80 eps=0.01                 Accuracy: \toriginal=0.590                 \tgroup:0.592                 \trandom:0.595                 \trandom_permute:0.095                 \trandom_g:0.590                 \trandom_inter_g:0.589                 Validation loss: \toriginal=1.869666                 \tgroup:1.892835                 \trandom:1.867456                 \trandom_permute:2.328465                 \trandom_g:1.869938                 \trandom_inter_g:1.869190\n","h=16 num_g=80 eps=0.001                 Accuracy: \toriginal=0.543                 \tgroup:0.544                 \trandom:0.545                 \trandom_permute:0.029                 \trandom_g:0.543                 \trandom_inter_g:0.543                 Validation loss: \toriginal=1.919117                 \tgroup:1.921114                 \trandom:1.919188                 \trandom_permute:2.401280                 \trandom_g:1.919100                 \trandom_inter_g:1.919073\n","h=16 num_g=80 eps=0.0001                 Accuracy: \toriginal=0.529                 \tgroup:0.529                 \trandom:0.529                 \trandom_permute:0.167                 \trandom_g:0.529                 \trandom_inter_g:0.529                 Validation loss: \toriginal=1.924030                 \tgroup:1.924110                 \trandom:1.924008                 \trandom_permute:2.289008                 \trandom_g:1.924028                 \trandom_inter_g:1.924013\n","h=16 num_g=100 eps=20.0                 Accuracy: \toriginal=0.492                 \tgroup:0.112                 \trandom:0.105                 \trandom_permute:0.112                 \trandom_g:0.057                 \trandom_inter_g:0.279                 Validation loss: \toriginal=1.967345                 \tgroup:nan                 \trandom:2.317021                 \trandom_permute:nan                 \trandom_g:2.307166                 \trandom_inter_g:2.187356\n","h=16 num_g=100 eps=10.0                 Accuracy: \toriginal=0.534                 \tgroup:0.292                 \trandom:0.112                 \trandom_permute:0.103                 \trandom_g:0.092                 \trandom_inter_g:0.351                 Validation loss: \toriginal=1.920583                 \tgroup:2.259248                 \trandom:2.310637                 \trandom_permute:2.309531                 \trandom_g:2.302660                 \trandom_inter_g:2.150268\n","h=16 num_g=100 eps=1.0                 Accuracy: \toriginal=0.559                 \tgroup:0.432                 \trandom:0.090                 \trandom_permute:0.106                 \trandom_g:0.129                 \trandom_inter_g:0.508                 Validation loss: \toriginal=1.901740                 \tgroup:2.223386                 \trandom:2.311085                 \trandom_permute:2.301896                 \trandom_g:2.291695                 \trandom_inter_g:2.097076\n","h=16 num_g=100 eps=0.1                 Accuracy: \toriginal=0.587                 \tgroup:0.588                 \trandom:0.552                 \trandom_permute:0.043                 \trandom_g:0.592                 \trandom_inter_g:0.585                 Validation loss: \toriginal=1.870037                 \tgroup:2.038532                 \trandom:1.943556                 \trandom_permute:2.321411                 \trandom_g:1.905002                 \trandom_inter_g:1.883671\n","h=16 num_g=100 eps=0.01                 Accuracy: \toriginal=0.565                 \tgroup:0.562                 \trandom:0.567                 \trandom_permute:0.078                 \trandom_g:0.566                 \trandom_inter_g:0.565                 Validation loss: \toriginal=1.888935                 \tgroup:1.907164                 \trandom:1.889018                 \trandom_permute:2.348304                 \trandom_g:1.887124                 \trandom_inter_g:1.887589\n","h=16 num_g=100 eps=0.001                 Accuracy: \toriginal=0.526                 \tgroup:0.525                 \trandom:0.524                 \trandom_permute:0.069                 \trandom_g:0.526                 \trandom_inter_g:0.526                 Validation loss: \toriginal=1.931776                 \tgroup:1.932061                 \trandom:1.932551                 \trandom_permute:2.379316                 \trandom_g:1.931759                 \trandom_inter_g:1.931882\n","h=16 num_g=100 eps=0.0001                 Accuracy: \toriginal=0.501                 \tgroup:0.501                 \trandom:0.501                 \trandom_permute:0.180                 \trandom_g:0.501                 \trandom_inter_g:0.501                 Validation loss: \toriginal=1.958536                 \tgroup:1.958601                 \trandom:1.958562                 \trandom_permute:2.270653                 \trandom_g:1.958539                 \trandom_inter_g:1.958545\n","h=16 num_g=120 eps=20.0                 Accuracy: \toriginal=0.570                 \tgroup:0.110                 \trandom:0.099                 \trandom_permute:0.110                 \trandom_g:0.079                 \trandom_inter_g:0.500                 Validation loss: \toriginal=1.886152                 \tgroup:nan                 \trandom:2.340575                 \trandom_permute:nan                 \trandom_g:2.303668                 \trandom_inter_g:2.201457\n","h=16 num_g=120 eps=10.0                 Accuracy: \toriginal=0.483                 \tgroup:0.280                 \trandom:0.110                 \trandom_permute:0.108                 \trandom_g:0.094                 \trandom_inter_g:0.343                 Validation loss: \toriginal=1.974414                 \tgroup:2.273032                 \trandom:2.310841                 \trandom_permute:2.302539                 \trandom_g:2.303687                 \trandom_inter_g:2.227090\n","h=16 num_g=120 eps=1.0                 Accuracy: \toriginal=0.577                 \tgroup:0.364                 \trandom:0.096                 \trandom_permute:0.091                 \trandom_g:0.112                 \trandom_inter_g:0.410                 Validation loss: \toriginal=1.884523                 \tgroup:2.218596                 \trandom:2.319038                 \trandom_permute:2.307963                 \trandom_g:2.301813                 \trandom_inter_g:2.092588\n","h=16 num_g=120 eps=0.1                 Accuracy: \toriginal=0.530                 \tgroup:0.533                 \trandom:0.438                 \trandom_permute:0.097                 \trandom_g:0.525                 \trandom_inter_g:0.518                 Validation loss: \toriginal=1.932140                 \tgroup:2.069854                 \trandom:2.039005                 \trandom_permute:2.305077                 \trandom_g:1.987187                 \trandom_inter_g:1.938328\n","h=16 num_g=120 eps=0.01                 Accuracy: \toriginal=0.509                 \tgroup:0.510                 \trandom:0.512                 \trandom_permute:0.059                 \trandom_g:0.511                 \trandom_inter_g:0.509                 Validation loss: \toriginal=1.946961                 \tgroup:1.959508                 \trandom:1.942151                 \trandom_permute:2.343342                 \trandom_g:1.945665                 \trandom_inter_g:1.946738\n","h=16 num_g=120 eps=0.001                 Accuracy: \toriginal=0.554                 \tgroup:0.554                 \trandom:0.554                 \trandom_permute:0.087                 \trandom_g:0.554                 \trandom_inter_g:0.554                 Validation loss: \toriginal=1.902542                 \tgroup:1.905683                 \trandom:1.902805                 \trandom_permute:2.361746                 \trandom_g:1.902607                 \trandom_inter_g:1.902683\n","h=16 num_g=120 eps=0.0001                 Accuracy: \toriginal=0.442                 \tgroup:0.442                 \trandom:0.442                 \trandom_permute:0.067                 \trandom_g:0.442                 \trandom_inter_g:0.442                 Validation loss: \toriginal=2.017788                 \tgroup:2.017738                 \trandom:2.017795                 \trandom_permute:2.392415                 \trandom_g:2.017788                 \trandom_inter_g:2.017793\n"]}]},{"cell_type":"code","source":["def plot(subplot_idx, acc_diff, vmax, title, cmap='coolwarm_r'):\n","    acc_diff = np.swapaxes(acc_diff, 0, 1)\n","    ax = plt.subplot(subplot_idx)\n","    im = ax.imshow(acc_diff, cmap=cmap, vmax=vmax, vmin=-vmax)\n","\n","    ax.set_xticks(np.arange(len(num_g_list)))\n","    ax.set_xticklabels(num_g_list)\n","    ax.set_yticks(np.arange(len(eps_list)))\n","    # ax.set_yticklabels(ticker.FormatStrFormatter('%.0e').format_ticks(eps_list))\n","    ax.set_yticklabels(eps_list)\n","    ax.set_xlabel('# of transformed models (num_g)',  fontsize=12)\n","    ax.set_ylabel(r'$\\epsilon$', fontsize=14)\n","\n","    plt.setp(ax.get_xticklabels(), rotation=45, ha=\"right\",\n","            rotation_mode=\"anchor\")\n","\n","    for num_g_idx in range(len(num_g_list)):\n","        for eps_idx in range(len(eps_list)):\n","            text = ax.text(eps_idx, num_g_idx, '{:.3g}'.format(acc_diff[num_g_idx, eps_idx]),\n","                          ha=\"center\", va=\"center\", color=\"w\", fontsize=6)\n","\n","    ax.set_title(title)\n","    return im\n","\n","\n","fig = plt.figure(figsize=(18,6))\n","max1 = np.max(abs(acc_group[0] - acc_original[0]))\n","max2 = np.max(abs(acc_random[0] - acc_original[0]))\n","max3 = np.max(abs(acc_random_permute[0] - acc_original[0]))\n","max4 = np.max(abs(acc_random_g[0] - acc_original[0]))\n","max5 = np.max(abs(acc_random_inter[0] - acc_original[0]))\n","vmax = np.max([max1, max2, max3, max4, max5])\n","# print('vmax acc: ', vmax)\n","plot(151, acc_group[0] - acc_original[0], vmax, 'accuracy (group - original)', cmap='coolwarm_r')\n","plot(152, acc_random_inter[0] - acc_original[0], vmax, 'accuracy (interpolated permute - original)', cmap='coolwarm_r')\n","plot(153, acc_random_g[0] - acc_original[0], vmax, 'accuracy (' + r'$g^{-1}$' + ' - original)', cmap='coolwarm_r')\n","plot(154, acc_random[0] - acc_original[0], vmax, 'accuracy (random - original)', cmap='coolwarm_r')\n","im = plot(155, acc_random_permute[0] - acc_original[0], vmax, 'accuracy (shuffle - original)', cmap='coolwarm_r')\n","\n","\n","plt.tight_layout()\n","fig.subplots_adjust(right=0.91)\n","cbar_ax = fig.add_axes([0.94, 0.18, 0.01, 0.6])\n","plt.colorbar(im, cax=cbar_ax)  \n","plt.savefig('figures/ensemble/cnn_cifar_leakyrelu2_main_acc.pdf')\n"],"metadata":{"id":"IzHxuAMtAY7r","executionInfo":{"status":"ok","timestamp":1677137590420,"user_tz":480,"elapsed":2897,"user":{"displayName":"Bo Zhao","userId":"08721035053344232673"}},"colab":{"base_uri":"https://localhost:8080/","height":290},"outputId":"fe890628-ef76-41c8-9375-76ecf0656aad"},"execution_count":10,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1296x432 with 6 Axes>"],"image/png":"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\n"},"metadata":{"image/png":{"width":1266,"height":273},"needs_background":"light"}}]},{"cell_type":"code","source":[],"metadata":{"id":"2mAtd_4FDtmP","executionInfo":{"status":"ok","timestamp":1677137590421,"user_tz":480,"elapsed":11,"user":{"displayName":"Bo Zhao","userId":"08721035053344232673"}}},"execution_count":10,"outputs":[]}]}