Some fixes and additional files

ais.py

@@ -14,7 +14,7 @@
 
 flags.DEFINE_string('datasource', 'random', 'default or noise or negative or single')
 flags.DEFINE_string('dataset', 'cifar10', 'cifar10 or mnist or dsprites or 2d or toy Gauss')
-flags.DEFINE_string('logdir', '/mnt/nfs/yilundu/ebm_code_release/cachedir', 'location where log of experiments will be stored')
+flags.DEFINE_string('logdir', 'cachedir', 'location where log of experiments will be stored')
 flags.DEFINE_string('exp', 'default', 'name of experiments')
 flags.DEFINE_integer('data_workers', 5, 'Number of different data workers to load data in parallel')
 flags.DEFINE_integer('batch_size', 16, 'Size of inputs')

ebm_combine.py

@@ -15,8 +15,8 @@
 
 flags.DEFINE_integer('batch_size', 256, 'Size of inputs')
 flags.DEFINE_integer('data_workers', 4, 'Number of workers to do things')
-flags.DEFINE_string('logdir', '/mnt/nfs/yilundu/ebm_code_release/cachedir', 'directory for logging')
-flags.DEFINE_string('savedir', '/mnt/nfs/yilundu/ebm_code_release/cachedir', 'location where log of experiments will be stored')
+flags.DEFINE_string('logdir', 'cachedir', 'directory for logging')
+flags.DEFINE_string('savedir', 'cachedir', 'location where log of experiments will be stored')
 flags.DEFINE_integer('num_filters', 64, 'number of filters for conv nets -- 32 for miniimagenet, 64 for omniglot.')
 flags.DEFINE_float('step_lr', 500, 'size of gradient descent size')
 flags.DEFINE_string('dsprites_path', '/root/data/dsprites-dataset/dsprites_ndarray_co1sh3sc6or40x32y32_64x64.npz', 'path to dsprites characters')
@@ -53,40 +53,6 @@
 
 FLAGS = flags.FLAGS
 
-if FLAGS.task == 'gentest' or FLAGS.task == 'genbaseline':
-    # For conditioning on large shapes
-    # FLAGS.exp_pos = "dpos_only_1_12_new"
-    # FLAGS.resume_pos = 3000
-
-    # For conditioning on small shapes that are orientated
-    # FLAGS.exp_pos = "dpsrites_dpos_only_107"
-    # FLAGS.resume_pos = 4000
-    # FLAGS.exp_size = 'dpsrites_dsize_only_107'
-    # FLAGS.resume_size = 1000
-
-    # For conditioning on small shapes that are not orientated
-    FLAGS.exp_pos = "dpos_only_1_16_replay_batch_2"
-    FLAGS.resume_pos = 4000
-
-    # Super strong model
-    # FLAGS.exp_size = "dscale_only_1_16_replay_batch"
-    # FLAGS.resume_size = 10000
-
-    FLAGS.exp_size = "dscale_only_1_17_replay_batch_2"
-    FLAGS.resume_size = 4000
-
-    # FLAGS.exp_size = "dscale_only_1_16_replay_batch"
-    # FLAGS.resume_size = 10000
-
-    FLAGS.exp_shape = "dshape_only_1_19_replay_batch"
-    FLAGS.resume_shape = 20000
-    FLAGS.exp_rot = "drot_only_1_16_replay_batch"
-    FLAGS.resume_rot = 7000
-    FLAGS.step_lr = 500
-    FLAGS.cond_shape = True
-    FLAGS.cond_rot = True
-
-
 class DSpritesGen(Dataset):
     def __init__(self, data, latents, frac=0.0):
 

ebm_sandbox.py

@@ -1,7 +1,6 @@
 import tensorflow as tf
 import math
 from tqdm import tqdm
-from hmc import hmc
 from tensorflow.python.platform import flags
 from torch.utils.data import DataLoader
 import torch
@@ -10,29 +9,28 @@
 from utils import optimistic_restore, set_seed
 import os.path as osp
 import numpy as np
-from rl_algs.logger import TensorBoardOutputFormat
+from baselines.logger import TensorBoardOutputFormat
 from scipy.misc import imsave
 import os
 import sklearn.metrics as sk
 from baselines.common.tf_util import initialize
 from scipy.linalg import eig
-from sklearn.manifold import TSNE
 import matplotlib.pyplot as plt
 
-set_seed(1)
+# set_seed(1)
 
 flags.DEFINE_string('datasource', 'random', 'default or noise or negative or single')
 flags.DEFINE_string('dataset', 'cifar10', 'omniglot or imagenet or omniglotfull or cifar10 or mnist or dsprites')
-flags.DEFINE_string('logdir', '/mnt/nfs/yilundu/ebm_code_release/sandbox_cachedir', 'location where log of experiments will be stored')
-flags.DEFINE_string('task', 'label', 'the task to execute (label: training on the label, anticorrupt: restore salt and pepper noise), boxcorrupt: restore empty portion of image'
+flags.DEFINE_string('logdir', 'sandbox_cachedir', 'location where log of experiments will be stored')
+flags.DEFINE_string('task', 'label', 'using conditional energy based models for classification'
+                    'anticorrupt: restore salt and pepper noise),'
+                    ' boxcorrupt: restore empty portion of image'
                     'or crossclass: change images from one class to another'
                     'or cycleclass: view image change across a label'
                     'or nearestneighbor which returns the nearest images in the test set'
-                    'or labelfinetune to train a model accuracy'
-                    'or latent to traverse the latent through energy')
-
+                    'or latent to traverse the latent space of an EBM through eigenvectors of the hessian (dsprites only)'
+                    'or mixenergy to evaluate out of distribution generalization compared to other datasets')
 flags.DEFINE_bool('hessian', True, 'Whether to use the hessian or the Jacobian for latent traversals')
-
 flags.DEFINE_string('exp', 'default', 'name of experiments')
 flags.DEFINE_integer('data_workers', 5, 'Number of different data workers to load data in parallel')
 flags.DEFINE_integer('batch_size', 32, 'Size of inputs')
@@ -45,22 +43,25 @@
 flags.DEFINE_bool('single', False, 'whether to use one sample to debug')
 flags.DEFINE_bool('cclass', True, 'whether to use a conditional model (required for task label)')
 flags.DEFINE_integer('num_steps', 20, 'number of steps to optimize the label')
-flags.DEFINE_integer('pgd', 0, 'number of steps project gradient descent to run')
-flags.DEFINE_integer('lnorm', -1, 'lnorm infinity is -1, ')
 flags.DEFINE_float('step_lr', 10.0, 'step size for updates on label')
 flags.DEFINE_float('proj_norm', 0.0, 'Maximum change of input images')
 flags.DEFINE_bool('large_model', False, 'Whether to use a large model')
-flags.DEFINE_bool('larger_model', False, 'Whether to use a large model')
-flags.DEFINE_bool('wider_model', False, 'Whether to use a large model')
+flags.DEFINE_bool('larger_model', False, 'Whether to use a larger model')
+flags.DEFINE_bool('wider_model', False, 'Whether to use a widermodel model')
 flags.DEFINE_bool('svhn', False, 'Whether to test on SVHN')
+
+# Conditions for mixenergy (outlier detection)
 flags.DEFINE_bool('svhnmix', False, 'Whether to test mix on SVHN')
 flags.DEFINE_bool('cifar100mix', False, 'Whether to test mix on CIFAR100')
-flags.DEFINE_bool('texturemix', False, 'Whether to test mix on CIFAR100')
-flags.DEFINE_bool('randommix', False, 'Whether to test mix on CIFAR100')
-flags.DEFINE_bool('groupsort', False, 'Whether to test mix on CIFAR100')
-flags.DEFINE_bool('hmc', False, 'Use HMC for cross class sampling')
+flags.DEFINE_bool('texturemix', False, 'Whether to test mix on Textures dataset')
+flags.DEFINE_bool('randommix', False, 'Whether to test mix on random dataset')
+
+# Conditions for label task (adversarial classification)
+flags.DEFINE_integer('lival', 8, 'Value of constraint for li')
+flags.DEFINE_integer('l2val', 40, 'Value of constraint for l2')
+flags.DEFINE_integer('pgd', 0, 'number of steps project gradient descent to run')
+flags.DEFINE_integer('lnorm', -1, 'linfinity is -1, l2 norm is 2')
 flags.DEFINE_bool('labelgrid', False, 'Make a grid of labels')
-flags.DEFINE_bool('proj_cclass', False, 'Projection conditional')
 
 # Conditions on which models to use
 flags.DEFINE_bool('cond_pos', True, 'whether to condition on position')
@@ -342,15 +343,23 @@ def boxcorrupt(test_dataloader, dataloader, weights, model, target_vars, logdir,
 
 
 def crossclass(dataloader, weights, model, target_vars, logdir, sess):
-    X, Y_GT, X_mods = target_vars['X'], target_vars['Y_GT'], target_vars['X_mods']
+    X, Y_GT, X_mods, X_final = target_vars['X'], target_vars['Y_GT'], target_vars['X_mods'], target_vars['X_final']
     for data_corrupt, data, label_gt in tqdm(dataloader):
         data, label_gt = data.numpy(), label_gt.numpy()
         data_corrupt = data.copy()
         data_corrupt[1:] = data_corrupt[0:-1]
         data_corrupt[0] = data[-1]
 
-        feed_dict = {X: data_corrupt, Y_GT: label_gt}
-        data_mods = sess.run(X_mods, feed_dict)
+        data_mods = []
+        data_mod = data_corrupt
+
+        for i in range(10):
+            data_mods.append(data_mod)
+
+            feed_dict = {X: data_mod, Y_GT: label_gt}
+            data_mod = sess.run(X_final, feed_dict)
+
+
 
         data_corrupt, data = rescale_im(data_corrupt), rescale_im(data)
 
@@ -798,33 +807,6 @@ def construct_steps(weights, X, Y_GT, model, target_vars):
     target_vars['X_mods'] = X_mods
 
 
-def construct_hmc_steps(weights, X, Y_GT, model, target_vars):
-    n = 50
-    scale_fac = 1.0
-
-    # if FLAGS.task == 'cycleclass':
-    #     scale_fac = 10.0
-
-    X_mods = []
-    X = tf.identity(X)
-
-    for i in range(FLAGS.num_steps):
-        p = tf.random_normal(tf.shape(X), mean=0.0, stddev=0.0001)
-        for j in range(10):
-            energy_noise = model.forward(X, weights, label=Y_GT, reuse=True)
-            x_grad = tf.gradients(energy_noise, [X])[0]
-            p = p - FLAGS.step_lr * x_grad / 2
-            X = X - FLAGS.step_lr * p
-
-            if i % n == (n-1):
-                X_mods.append(X)
-
-            print("Constructing step {}".format(i))
-
-    target_vars['X_final'] = X
-    target_vars['X_mods'] = X_mods
-
-
 def nearest_neighbor(dataset, sess, target_vars, logdir):
     X = target_vars['X']
     Y_GT = target_vars['Y_GT']
@@ -886,7 +868,10 @@ def main():
     elif FLAGS.larger_model:
         model = ResNet32Larger(num_filters=hidden_dim)
     elif FLAGS.wider_model:
-        model = ResNet32Wider(num_filters=196, train=False)
+        if FLAGS.dataset == 'imagenet':
+            model = ResNet32Wider(num_filters=196, train=False)
+        else:
+            model = ResNet32Wider(num_filters=256, train=False)
     else:
         model = ResNet32(num_filters=hidden_dim)
 
@@ -925,20 +910,17 @@ def main():
     if FLAGS.task == 'label':
         construct_label(weights, X, Y, Y_GT, model, target_vars)
     elif FLAGS.task == 'labelfinetune':
-        construct_finetune_label(weights, X, Y, Y_GT, model, target_vars)
+        construct_finetune_label(weights, X, Y, Y_GT, model, target_vars, )
     elif FLAGS.task == 'energyeval' or FLAGS.task == 'mixenergy':
         construct_energy(weights, X, Y, Y_GT, model, target_vars)
     elif FLAGS.task == 'anticorrupt' or FLAGS.task == 'boxcorrupt' or FLAGS.task == 'crossclass' or FLAGS.task == 'cycleclass' or FLAGS.task == 'democlass' or FLAGS.task == 'nearestneighbor':
-        if FLAGS.hmc:
-            construct_hmc_steps(weights, X, Y_GT, model, target_vars)
-        else:
-            construct_steps(weights, X, Y_GT, model, target_vars)
+        construct_steps(weights, X, Y_GT, model, target_vars)
     elif FLAGS.task == 'latent':
         construct_latent(weights, X, Y_GT, model, target_vars)
 
     sess.run(tf.global_variables_initializer())
     saver = loader = tf.train.Saver(max_to_keep=10)
-    savedir = osp.join('/mnt/nfs/yilundu/ebm_code_release/cachedir', FLAGS.exp)
+    savedir = osp.join('cachedir', FLAGS.exp)
     logdir = osp.join(FLAGS.logdir, FLAGS.exp)
     if not osp.exists(logdir):
         os.makedirs(logdir)
@@ -948,7 +930,7 @@ def main():
         model_file = osp.join(savedir, 'model_{}'.format(FLAGS.resume_iter))
         resume_itr = FLAGS.resume_iter
 
-        if FLAGS.task == 'label' or FLAGS.task == 'boxcorrupt' or FLAGS.task == 'labelfinetune' or FLAGS.task == "energyeval":
+        if FLAGS.task == 'label' or FLAGS.task == 'boxcorrupt' or FLAGS.task == 'labelfinetune' or FLAGS.task == "energyeval" or FLAGS.task == "crossclass" or FLAGS.task == "mixenergy":
             optimistic_restore(sess, model_file)
             # saver.restore(sess, model_file)
         else:
@@ -971,7 +953,7 @@ def main():
         else:
             label(dataloader, test_dataloader, target_vars, sess)
     elif FLAGS.task == 'labelfinetune':
-        labelfinetune(dataloader, test_dataloader, target_vars, sess, savedir, saver)
+        labelfinetune(dataloader, test_dataloader, target_vars, sess, savedir, saver, l1val=FLAGS.lival, l2val=FLAGS.l2val)
     elif FLAGS.task == 'energyeval':
         energyeval(dataloader, test_dataloader, target_vars, sess)
     elif FLAGS.task == 'mixenergy':

hmc.py

@@ -97,8 +97,8 @@ def hmc(initial_x,
         Step-size in Hamiltonian simulation
     num_steps : int
         Number of steps to take in Hamiltonian simulation
-    log_posterior : str
-        Log posterior (unnormalized) for the target distribution
+    neg_log_posterior : str
+        Negative log posterior (unnormalized) for the target distribution
 
     Returns
     -------

imagenet_demo.py

@@ -6,7 +6,7 @@
 import imageio
 
 
-flags.DEFINE_string('logdir', '/mnt/nfs/yilundu/ebm_code_release/cachedir', 'location where log of experiments will be stored')
+flags.DEFINE_string('logdir', '../cachedir', 'location where log of experiments will be stored')
 flags.DEFINE_integer('num_steps', 200, 'num of steps for conditional imagenet sampling')
 flags.DEFINE_float('step_lr', 170., 'number of steps to run')
 flags.DEFINE_integer('batch_size', 16, 'number of steps to run')
@@ -63,9 +63,10 @@ def rescale_im(im):
     model_file = osp.join(logdir, 'model_{}'.format(FLAGS.resume_iter))
     saver.restore(sess, model_file)
 
-    # lx = list(range(16))
-    lx = [0, 14, 37, 145, 108, 963, 242, 624, 238, 323, 527, 748, 985, 973, 974, 979]
+    ls = np.random.permutation(1000)[:16]
     ims = []
+
+    # What to initialize sampling with. 
     x_mod = np.random.uniform(0, 1, size=(FLAGS.batch_size, 128, 128, 3))
     labels = np.eye(1000)[lx]
 

requirements.txt

@@ -0,0 +1,16 @@
+scipy==0.19.1
+horovod==0.16.0
+torch==0.3.1
+scikit_image==0.13.0
+tensorflow==1.12.0
+torchvision==0.2.0
+six==1.11.0
+imageio==2.4.1
+tqdm==4.20.0
+matplotlib==1.5.3
+mpi4py==3.0.0
+numpy==1.14.0
+Pillow==5.4.1
+baselines==0.1.5
+skimage==0.0
+scikit_learn==0.20.3

test_inception.py

@@ -18,7 +18,7 @@
 from inception import get_inception_score
 from fid import get_fid_score
 
-flags.DEFINE_string('logdir', '/mnt/nfs/yilundu/ebm_code_release/cachedir', 'location where log of experiments will be stored')
+flags.DEFINE_string('logdir', 'cachedir', 'location where log of experiments will be stored')
 flags.DEFINE_string('exp', 'default', 'name of experiments')
 flags.DEFINE_bool('cclass', False, 'whether to condition on class')
 
@@ -216,7 +216,7 @@ def compute_inception(sess, target_vars):
 
         images.extend(list(ims))
 
-    saveim = osp.join('/mnt/nfs/yilundu/ebm_code_release/sandbox_cachedir', FLAGS.exp, "test{}.png".format(FLAGS.idx))
+    saveim = osp.join('sandbox_cachedir', FLAGS.exp, "test{}.png".format(FLAGS.idx))
 
     ims = ims[:100]
 

train.py

@@ -48,12 +48,12 @@
     'Number of different data workers to load data in parallel')
 
 # General Experiment Settings
-flags.DEFINE_string('logdir', '/mnt/nfs/yilundu/ebm_code_release/cachedir',
+flags.DEFINE_string('logdir', 'cachedir',
     'location where log of experiments will be stored')
 flags.DEFINE_string('exp', 'default', 'name of experiments')
 flags.DEFINE_integer('log_interval', 10, 'log outputs every so many batches')
-flags.DEFINE_integer('save_interval', 2000,'save outputs every so many batches')
-flags.DEFINE_integer('test_interval', 2000,'evaluate outputs every so many batches')
+flags.DEFINE_integer('save_interval', 1000,'save outputs every so many batches')
+flags.DEFINE_integer('test_interval', 1000,'evaluate outputs every so many batches')
 flags.DEFINE_integer('resume_iter', -1, 'iteration to resume training from')
 flags.DEFINE_bool('train', True, 'whether to train or test')
 flags.DEFINE_integer('epoch_num', 10000, 'Number of Epochs to train on')