class GaussianMixture(Distribution):
def __init__(self, normal_means, normal_stds, weights):
self.num_gaussians = weights.shape[1]
self.normal_means = normal_means
self.normal_stds = normal_stds
self.normal = MultivariateDiagonalNormal(normal_means, normal_stds)
self.normals = [
MultivariateDiagonalNormal(normal_means[:, :, i], normal_stds[:, :,
i])
for i in range(self.num_gaussians)
]
self.weights = weights
self.categorical = OneHotCategorical(self.weights[:, :, 0])
def log_prob(
self,
value,
):
log_p = [
self.normals[i].log_prob(value) for i in range(self.num_gaussians)
]
log_p = torch.stack(log_p, -1)
log_p = log_p.sum(dim=1)
log_weights = torch.log(self.weights[:, :, 0])
lp = log_weights + log_p
m = lp.max(dim=1)[0] # log-sum-exp numerical stability trick
log_p_mixture = m + torch.log(torch.exp(lp - m).sum(dim=1))
return log_p_mixture
def sample(self):
z = self.normal.sample().detach()
c = self.categorical.sample()[:, :, None]
s = torch.matmul(z, c)
return torch.squeeze(s, 2)
def rsample(self):
z = (self.normal_means + self.normal_stds * MultivariateDiagonalNormal(
ptu.zeros(self.normal_means.size()),
ptu.ones(self.normal_stds.size())).sample())
z.requires_grad_()
c = self.categorical.sample()[:, :, None]
s = torch.matmul(z, c)
return torch.squeeze(s, 2)
def mle_estimate(self):
"""Return the mean of the most likely component.
This often computes the mode of the distribution, but not always.
"""
c = ptu.zeros(self.weights.shape[:2])
ind = torch.argmax(self.weights, dim=1) # [:, 0]
c.scatter_(1, ind, 1)
s = torch.matmul(self.normal_means, c[:, :, None])
return torch.squeeze(s, 2)
def __repr__(self):
s = "GaussianMixture(normal_means=%s, normal_stds=%s, weights=%s)"
return s % (self.normal_means, self.normal_stds, self.weights)
class GaussianMixture(Distribution):
def __init__(self, normal_means, normal_stds, weights):
self.num_gaussians = weights.shape[1]
self.normal_means = normal_means
self.normal_stds = normal_stds
self.normal = Normal(normal_means, normal_stds)
self.normals = [
Normal(normal_means[:, :, i], normal_stds[:, :, i])
for i in range(self.num_gaussians)
]
self.weights = weights
self.categorical = OneHotCategorical(self.weights[:, :, 0])
def log_prob(
self,
value,
):
log_p = [
self.normals[i].log_prob(value) for i in range(self.num_gaussians)
]
log_p = torch.stack(log_p, -1)
log_p = log_p.sum(dim=1)
log_weights = torch.log(self.weights[:, :, 0])
lp = log_weights + log_p
m = lp.max(dim=1,
keepdim=True)[0] # log-sum-exp numerical stability trick
log_p_mixture = m + torch.log(
torch.exp(lp - m).sum(dim=1, keepdim=True))
return log_p_mixture
def sample(self, ):
z = self.normal.sample().detach()
c = self.categorical.sample()[:, :, None]
s = torch.matmul(z, c)
return torch.squeeze(s, 2)
def rsample(self, ):
z = (self.normal_means + self.normal_stds *
Normal(ptu.zeros(self.normal_means.size()),
ptu.ones(self.normal_stds.size())).sample())
z.requires_grad_()
c = self.categorical.sample()[:, :, None]
s = torch.matmul(z, c)
return torch.squeeze(s, 2)
def mean(self, ):
"""Misleading function name; this actually now samples the mean of the
most likely component.
c ~ argmax(C), returns mu_c
This often computes the mode of the distribution, but not always.
"""
c = ptu.zeros(self.weights.shape[:2])
ind = torch.argmax(self.weights, dim=1) # [:, 0]
c.scatter_(1, ind, 1)
s = torch.matmul(self.normal_means, c[:, :, None])
return torch.squeeze(s, 2)
def test_one_hot_categorical_shape(self):
dist = OneHotCategorical(torch.Tensor([[0.6, 0.3], [0.6, 0.3], [0.6, 0.3]]))
self.assertEqual(dist._batch_shape, torch.Size((3,)))
self.assertEqual(dist._event_shape, torch.Size((2,)))
self.assertEqual(dist.sample().size(), torch.Size((3, 2)))
self.assertEqual(dist.sample((3, 2)).size(), torch.Size((3, 2, 3, 2)))
self.assertEqual(dist.log_prob(self.tensor_sample_1).size(), torch.Size((3,)))
self.assertRaises(ValueError, dist.log_prob, self.tensor_sample_2)
self.assertEqual(dist.log_prob(dist.enumerate_support()).size(), torch.Size((2, 3)))
class TestMultiOneHotCategorical(unittest.TestCase):
def setUp(self) -> None:
self.test_probs = torch.tensor([[0.3, 0.2, 0.4, 0.1, 0.25, 0.5, 0.25, 0.3, 0.4, 0.1, 0.1, 0.1],
[0.2, 0.3, 0.1, 0.4, 0.5, 0.3, 0.2, 0.2, 0.3, 0.2, 0.2, 0.1]])
self.test_sections = (4, 3, 5)
self.test_actions = torch.tensor([[0., 0., 1., 0., 0., 1., 0., 0., 1., 0., 0., 0.],
[0., 0., 0., 1., 1., 0., 0., 0., 0., 1., 0., 0.]]).long()
self.test_sected_actions = torch.split(self.test_actions, self.test_sections, dim=-1)
self.test_multi_onehot_categorical = MultiOneHotCategorical(self.test_probs, self.test_sections)
self.test_onehot_categorical1 = OneHotCategorical(self.test_probs[:, :4])
self.test_onehot_categorical2 = OneHotCategorical(self.test_probs[:, 4:7])
self.test_onehot_categorical3 = OneHotCategorical(self.test_probs[:, 7:])
def test_log_prob(self):
test_cat1_log_prob = self.test_onehot_categorical1.log_prob(self.test_sected_actions[0])
test_cat2_log_prob = self.test_onehot_categorical2.log_prob(self.test_sected_actions[1])
test_cat3_log_prob = self.test_onehot_categorical3.log_prob(self.test_sected_actions[2])
test_multi_cat_log_prob = self.test_multi_onehot_categorical.log_prob(self.test_actions)
print(test_multi_cat_log_prob)
print(test_cat1_log_prob)
self.assertEqual(test_cat1_log_prob.shape, test_multi_cat_log_prob.shape)
self.assertTrue(
torch.equal(test_cat1_log_prob + test_cat2_log_prob + test_cat3_log_prob, test_multi_cat_log_prob))
def test_sample(self):
test_cat1_sample = self.test_onehot_categorical1.sample()
test_cat2_sample = self.test_onehot_categorical2.sample()
test_cat3_sample = self.test_onehot_categorical3.sample()
test_cat_sample = torch.cat([test_cat1_sample, test_cat2_sample, test_cat3_sample], dim=-1)
test_multi_cat_sample = self.test_multi_onehot_categorical.sample()
self.assertEqual(test_cat_sample.shape, test_multi_cat_sample.shape)
self.assertTrue(torch.equal(test_cat_sample.sum(dim=-1),
test_multi_cat_sample.sum(dim=-1)))
def test_entropy(self):
test_cat1_entropy = self.test_onehot_categorical1.entropy()
test_cat2_entropy = self.test_onehot_categorical2.entropy()
test_cat3_entropy = self.test_onehot_categorical3.entropy()
test_multi_cat_entropy = self.test_multi_onehot_categorical.entropy()
self.assertTrue(torch.equal(test_cat1_entropy + test_cat2_entropy + test_cat3_entropy, test_multi_cat_entropy),
"Expected same entropy!!!")
def forward(self,
inp: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor, None]:
"""
Returns a sample of the policy on the input with the mean and log
probability of the sample.
Args:
inp: The input tensor to put through the network.categorie
Returns:
A multi-categorical distribution of the network.
"""
linear = self.linear(inp)
value = self.value(linear)
value = value.view(*value.shape[:-1], -1, self.num_classes)
probs = self.probs(value)
dist = OneHotCategorical(probs)
sample = dist.sample()
log_prob = dist.log_prob(sample)
# Straight through gradient trick
sample = sample + probs - probs.detach()
mean = torch.argmax(probs, dim=-1).values
return sample, log_prob, None
def forward(self, x):
# For convenience we use torch.distributions to sample and compute the values of interest for the distribution see (https://pytorch.org/docs/stable/distributions.html) for more details.
probs = self.encode(x.view(-1, 784))
m = OneHotCategorical(probs)
action = m.sample()
log_prob = m.log_prob(action)
entropy = m.entropy()
return self.decode(action), log_prob, entropy
def rsample(self, sample_shape=torch.Size()):
if len(sample_shape) > 0:
cat = OneHotCategorical(logits=self.logits).sample(
sample_shape=sample_shape)
else:
cat = OneHotCategorical(logits=self.logits.transpose(1, -1))
cat = cat.sample(sample_shape=sample_shape).transpose(1, -1)
cat = cat[:, :, None]
print("LS", self.locs.shape, cat.shape, self.logits.shape)
loc = (self.locs * cat).sum(dim=1)
scale = (self.scales * cat).sum(dim=1)
dist = Normal(loc, scale)
return dist.sample()
def test_one_hot_categorical_2d(self):
probabilities = [[0.1, 0.2, 0.3], [0.5, 0.3, 0.2]]
probabilities_1 = [[1.0, 0.0], [0.0, 1.0]]
p = Variable(torch.Tensor(probabilities), requires_grad=True)
s = Variable(torch.Tensor(probabilities_1), requires_grad=True)
self.assertEqual(OneHotCategorical(p).sample().size(), (2, 3))
self.assertEqual(OneHotCategorical(p).sample(sample_shape=(3, 4)).size(), (3, 4, 2, 3))
self.assertEqual(OneHotCategorical(p).sample_n(6).size(), (6, 2, 3))
self._gradcheck_log_prob(OneHotCategorical, (p,))
dist = OneHotCategorical(p)
x = dist.sample()
self.assertEqual(dist.log_prob(x), Categorical(p).log_prob(x.max(-1)[1]))
def sample_from_density(cls, n_sample: int, response_net: nn.Module,
norm: Normal, cat: OneHotCategorical,
covariate: Optional[torch.Tensor]):
pred_list = []
for i in range(n_sample):
cat_sample = cat.sample().unsqueeze(
1) # size = [n_data, 1, n_component]
norm_sample = norm.sample(
) # size = [n_data, output_dim, n_component]
sample = torch.sum(cat_sample * norm_sample,
dim=2) # size = [n_data, output_dim]
if covariate is not None:
pred = response_net(sample, covariate)
else:
pred = response_net(sample)
pred_list.append(pred)
return torch.cat(pred_list, dim=0)
def forward(self, input, args, n_particles, test=False):
"""
evaluation is the IWAE-10 bound
"""
pi = F.log_softmax(self.pi, 0)
# run the input and teacher-forcing inputs through the embedding layers here
seq_len, batch_sz = input.size()
emb = self.inp_embedding(input)
hidden = self.init_hidden(batch_sz, self.nhid, 2) # bidirectional
hidden_states, (_, _) = self.encoder(emb, hidden)
hidden_states = hidden_states.repeat(1, n_particles, 1)
# run the z-decoder at this point, evaluating the NLL at each step
h = (Variable(
hidden_states.data.new(batch_sz * n_particles,
self.hidden_size).zero_()),
Variable(
hidden_states.data.new(batch_sz * n_particles,
self.hidden_size).zero_()))
nlls = hidden_states.data.new(seq_len, batch_sz * n_particles)
loss = 0
# now a log-prob
prior_logits = pi.unsqueeze(0).expand(batch_sz * n_particles,
self.z_dim)
prior_h = (Variable(torch.zeros(batch_sz * n_particles, 50).cuda()),
Variable(torch.zeros(batch_sz * n_particles, 50).cuda()))
accumulated_weights = -math.log(
n_particles) # will contain log w_{t - 1}
logits = self.init_hidden(batch_sz * n_particles,
self.z_dim,
squeeze=True)
feed = None
x_emb = self.lockdrop(emb, self.dropout_x)
if test:
pdb.set_trace()
for i in range(seq_len):
# build the next z sample - not differentiable! we don't train the inference network
logits = F.log_softmax(self.logits(hidden_states[i]), 1).detach()
# if test:
q = OneHotCategorical(logits=logits)
p = OneHotCategorical(logits=prior_logits)
a = q.sample()
# else:
# q = RelaxedOneHotCategorical(temperature=self.temp, logits=logits)
# p = RelaxedOneHotCategorical(temperature=self.temp_prior, logits=prior_logits)
# a = q.rsample()
# to guard against being too crazy
b = a + 1e-16
z = b / b.sum(1, keepdim=True)
# this should be batch_sz x x_dim
scores = torch.mm(self.project(torch.cat([h[0], z], 1)),
self.emit.t())
NLL = nn.CrossEntropyLoss(reduce=False)(
scores, input[i].repeat(n_particles))
nlls[i] = NLL.data
f_term = p.log_prob(z) # prior
r_term = q.log_prob(z) # proposal
alpha = -NLL + (f_term - r_term)
wa = accumulated_weights + alpha.view(n_particles, batch_sz)
Z = log_sum_exp(wa, dim=0) # line 7
loss += Z # line 8
accumulated_weights = wa - Z # F.log_softmax(wa, dim=0) # line 9
probs = accumulated_weights.data.exp()
probs += 0.01
probs = probs / probs.sum(0, keepdim=True)
effective_sample_size = 1. / probs.pow(2).sum(0)
if any_nans(probs):
pdb.set_trace()
# probs is [n_particles, batch_sz]
# ancestors [2 x 15] = [[0, 0, 0, ..., 0], [0, 1, 2, 3, ...]]
# offsets [2 x 15] = [[0, 0, 0, ..., 0], [1, 1, 1, 1, ...]]
# resample / RSAMP
if ((effective_sample_size / n_particles) < 0.3).sum() > 0:
ancestors = torch.multinomial(probs.transpose(0, 1),
n_particles, True)
# now, reindex, which is the most important thing
offsets = n_particles * torch.arange(batch_sz).unsqueeze(
1).repeat(1, n_particles).long()
if ancestors.is_cuda:
offsets = offsets.cuda()
unrolled_idx = Variable(ancestors + offsets).view(-1)
# shuffle!
z = torch.index_select(z, 0, unrolled_idx)
a, b = h
h = torch.index_select(a, 0, unrolled_idx), torch.index_select(
b, 0, unrolled_idx)
a, b = prior_h
prior_h = torch.index_select(a, 0,
unrolled_idx), torch.index_select(
b, 0, unrolled_idx)
# reset accumulated_weights
accumulated_weights = -math.log(
n_particles) # will contain log w_{t - 1}
# set things up for next time
if i != seq_len - 1:
feed = torch.cat(
[emb[i].repeat(n_particles, 1),
self.z_emb(z)], 1)
prior_h = self.z_decoder(feed, prior_h)
prior_logits = F.log_softmax(self.project_z(prior_h[0]), 1)
h = self.hidden_rnn(x_emb[i].repeat(n_particles, 1),
h) # feed the next word into the RNN
NLL = nlls
# now, we calculate the final log-marginal estimator
return loss.sum(), NLL.sum(), (seq_len * batch_sz * n_particles), 0
nn.Linear(hidden, z_num))
def forward(self, s, log=False):
feature = self.net(s)
if log:
return F.log_softmax(feature, dim=-1)
else:
return F.softmax(feature, dim=-1)
if __name__ == "__main__":
from torch.distributions import Categorical
from torch.distributions import OneHotCategorical
onehot = OneHotCategorical(torch.ones(4))
s = torch.FloatTensor([1, 2])
z = onehot.sample() #torch.LongTensor([0,0,1,0])
print(s, z)
policy = Policy(s_dim=2, z_num=4, hidden=32, a_num=4)
vnet = VNet(s_dim=2, z_num=4, hidden=32)
qnet = QNet(s_dim=2, z_num=4, hidden=32, a_num=4)
dis = Discriminator(s_dim=2, z_num=4, hidden=32)
prob = policy(s, z)
print(prob)
dist = Categorical(prob)
a = dist.sample()
print(a)
index = torch.LongTensor(range(1))
v = vnet(s, z)
q = qnet(s, z)
print(v, q)
def forward(self, input, targets, args, n_particles, criterion, test=False):
"""
This version takes the inputs, and does not expose the logits, but instead
computes the losses directly
"""
# run the input and teacher-forcing inputs through the embedding layers here
seq_len, batch_sz = input.size()
emb = self.inp_embedding(input)
hidden = self.init_hidden(batch_sz, self.nhid, 2) # bidirectional
hidden_states, (h, c) = self.encoder(emb, hidden)
# teacher-forcing
out_emb = self.dropout(self.dec_embedding(targets))
# now, we'll replicate it for each particle - it's currently [seq_len x batch_sz x nhid]
hidden_states = hidden_states.repeat(1, n_particles, 1)
out_emb = out_emb.repeat(1, n_particles, 1)
# now [seq_len x (n_particles x batch_sz) x nhid]
# out_emb, hidden_states should be viewed as (n_particles x batch_sz) - this means that's true for h as well
# run the z-decoder at this point, evaluating the NLL at each step
p_h = self.init_hidden(batch_sz * n_particles, self.z_dim, squeeze=True) # initially zero
h = self.init_hidden(batch_sz * n_particles, self.z_dim, squeeze=True)
d_h = self.init_hidden(batch_sz * n_particles, self.nhid, squeeze=True)
nlls = hidden_states.data.new(seq_len, batch_sz * n_particles)
loss = 0
accumulated_weights = -math.log(n_particles) # will contain log w_{t - 1}
resamples = 0
for i in range(seq_len):
h = self.z_decoder(hidden_states[i], h)
logits = self.logits(h)
# build the next z sample
if test:
q = OneHotCategorical(logits=logits)
z = q.sample()
else:
q = RelaxedOneHotCategorical(temperature=self.temp, logits=logits)
z = q.rsample()
h = z
# prior
if test:
p = OneHotCategorical(logits=p_h)
else:
p = RelaxedOneHotCategorical(temperature=self.temp_prior, logits=p_h)
# now, compute the log-likelihood of the data given this mean, and the input out_emb
d_h = self.decoder(torch.cat([z, out_emb[i]], 1), d_h)
decoder_logits = self.out_embedding(d_h)
NLL = criterion(decoder_logits, input[i].repeat(n_particles))
nlls[i] = NLL.data
# compute the weight using `reweight` on page (4)
f_term = p.log_prob(z) # prior
r_term = q.log_prob(z) # proposal
alpha = -NLL + args.anneal * (f_term - r_term)
wa = accumulated_weights + alpha.view(n_particles, batch_sz)
# sample ancestors, and reindex everything
Z = log_sum_exp(wa, dim=0) # line 7
if (Z.data > 0.1).any():
pdb.set_trace()
loss += Z # line 8
accumulated_weights = wa - Z # line 9
probs = accumulated_weights.data.exp()
probs += 0.01
probs = probs / probs.sum(0, keepdim=True)
effective_sample_size = 1./probs.pow(2).sum(0)
# resample / RSAMP if 3 batch elements need resampling
if ((effective_sample_size / n_particles) < 0.3).sum() > 0:
resamples += 1
ancestors = torch.multinomial(probs.transpose(0, 1), n_particles, True)
# now, reindex, which is the most important thing
offsets = n_particles * torch.arange(batch_sz).unsqueeze(1).repeat(1, n_particles).long()
if ancestors.is_cuda:
offsets = offsets.cuda()
unrolled_idx = Variable(ancestors.t().contiguous()+offsets).view(-1)
h = torch.index_select(h, 0, unrolled_idx)
p_h = torch.index_select(p_h, 0, unrolled_idx)
d_h = torch.index_select(d_h, 0, unrolled_idx)
# reset accumulated_weights
accumulated_weights = -math.log(n_particles) # will contain log w_{t - 1}
if i != seq_len - 1:
# build the next mean prediction, feeding in the correct ancestor
p_h = self.ar_prior_logits(torch.cat([h, out_emb[i]], 1), p_h)
# now, we calculate the final log-marginal estimator
nll = nlls.view(seq_len, n_particles, batch_sz).mean(1).sum()
return -loss.sum(), nll, (seq_len * batch_sz), resamples
def forward(self, input, args, n_particles, test=False):
T = F.log_softmax(self.T, 0) # NOTE: in log-space
pi = F.log_softmax(self.pi, 0) # NOTE: in log-space
emit = self.calc_emit()
# run the input and teacher-forcing inputs through the embedding layers here
seq_len, batch_sz = input.size()
emb = self.inp_embedding(input)
hidden = self.init_hidden(batch_sz, self.nhid, 2) # bidirectional
hidden_states, (_, _) = self.encoder(emb, hidden)
hidden_states = hidden_states.repeat(1, n_particles, 1)
# run the z-decoder at this point, evaluating the NLL at each step
h = self.init_hidden(batch_sz * n_particles, self.z_dim, squeeze=True)
nlls = hidden_states.data.new(seq_len, batch_sz * n_particles)
loss = 0
accumulated_weights = -math.log(
n_particles) # will contain log w_{t - 1}
resamples = 0
# in log probability space
prior_probs = pi.unsqueeze(0).expand(batch_sz * n_particles,
self.z_dim)
logits = self.init_hidden(batch_sz * n_particles,
self.z_dim,
squeeze=True)
for i in range(seq_len):
# logits = self.logits(nn.functional.relu(self.z_decoder(hidden_states[i], logits)))
logits = self.logits(
nn.functional.relu(
self.z_decoder(torch.cat([hidden_states[i], h], 1),
logits)))
# build the next z sample
if any_nans(logits):
pdb.set_trace()
if test:
q = OneHotCategorical(logits=logits)
z = q.sample()
else:
q = RelaxedOneHotCategorical(temperature=self.temp,
logits=logits)
z = q.rsample()
h = z
# prior
if any_nans(prior_probs):
pdb.set_trace()
if test:
p = OneHotCategorical(logits=prior_probs)
else:
p = RelaxedOneHotCategorical(temperature=self.temp_prior,
logits=prior_probs)
if any_nans(prior_probs):
pdb.set_trace()
if any_nans(logits):
pdb.set_trace()
# now, compute the log-likelihood of the data given this z-sample
NLL = -self.decode(z, input[i].repeat(n_particles),
(emit, )) # diff. w.r.t. z
nlls[i] = NLL.data
# compute the weight using `reweight` on page (4)
f_term = p.log_prob(z) # prior
r_term = q.log_prob(z) # proposal
alpha = -NLL + (f_term - r_term)
wa = accumulated_weights + alpha.view(n_particles, batch_sz)
# sample ancestors, and reindex everything
Z = log_sum_exp(wa, dim=0) # line 7
loss += Z # line 8
accumulated_weights = wa - Z # line 9
if args.filter:
probs = accumulated_weights.data.exp()
probs += 0.01
probs = probs / probs.sum(0, keepdim=True)
effective_sample_size = 1. / probs.pow(2).sum(0)
# probs is [n_particles, batch_sz]
# ancestors [2 x 15] = [[0, 0, 0, ..., 0], [0, 1, 2, 3, ...]]
# offsets [2 x 15] = [[0, 0, 0, ..., 0], [1, 1, 1, 1, ...]]
# resample / RSAMP
if ((effective_sample_size / n_particles) < 0.3).sum() > 0:
resamples += 1
ancestors = torch.multinomial(probs.transpose(0, 1),
n_particles, True)
# now, reindex, which is the most important thing
offsets = n_particles * torch.arange(batch_sz).unsqueeze(
1).repeat(1, n_particles).long()
if ancestors.is_cuda:
offsets = offsets.cuda()
unrolled_idx = Variable(ancestors + offsets).view(-1)
h = torch.index_select(h, 0, unrolled_idx)
# reset accumulated_weights
accumulated_weights = -math.log(
n_particles) # will contain log w_{t - 1}
if i != seq_len - 1:
# now in probability space
prior_probs = log_sum_exp(T.unsqueeze(0) + z.unsqueeze(1), 2)
# let's normalize things - slower, but safer
# prior_probs += 0.01
# prior_probs = prior_probs / prior_probs.sum(1, keepdim=True)
# # if ((prior_probs.sum(1) - 1) > 1e-3).any()[0]:
# pdb.set_trace()
if any_nans(loss):
pdb.set_trace()
# now, we calculate the final log-marginal estimator
return -loss.sum(), nlls.sum(), (seq_len * batch_sz *
n_particles), resamples
def test(epoch):
model.eval()
test_loss = 0
with torch.no_grad():
for i, (data, _) in enumerate(test_loader):
data = data.to(device)
recon_batch, log_prob, entropy = model(data)
test_loss += loss_function(recon_batch, data, log_prob, entropy).item()
if i == 0:
n = min(data.size(0), 8)
comparison = torch.cat([data[:n],
recon_batch.view(args.batch_size, 1, 28, 28)[:n]])
save_image(comparison.cpu(),
'results/reconstruction_' + str(epoch) + '.png', nrow=n)
test_loss /= len(test_loader.dataset)
print('====> Test set loss: {:.4f}'.format(test_loss))
for epoch in range(1, args.epochs + 1):
train(epoch)
test(epoch)
with torch.no_grad():
m = OneHotCategorical(torch.ones(256)/256.)
sample = m.sample((64, 20))
sample = sample.to(device)
sample = model.decode(sample).cpu()
save_image(sample.view(64, 1, 28, 28),
'results/sample_' + str(epoch) + '.png')
def forward(self, input, args, n_particles, test=False):
T = nn.Softmax(dim=0)(self.T) # NOTE: not in log-space
pi = nn.Softmax(dim=0)(self.pi)
emit = self.calc_emit()
# run the input and teacher-forcing inputs through the embedding layers here
seq_len, batch_sz = input.size()
emb = self.inp_embedding(input)
hidden = self.init_hidden(batch_sz, self.nhid, 2) # bidirectional
hidden_states, (_, _) = self.encoder(emb, hidden)
hidden_states = hidden_states.repeat(1, n_particles, 1)
# run the z-decoder at this point, evaluating the NLL at each step
z = self.init_hidden(batch_sz * n_particles, self.z_dim, squeeze=True)
nlls = hidden_states.data.new(seq_len, batch_sz * n_particles)
loss = 0
# now a log-prob
prior_probs = pi.unsqueeze(0).expand(batch_sz * n_particles,
self.z_dim)
logits = self.init_hidden(batch_sz * n_particles,
self.z_dim,
squeeze=True)
for i in range(seq_len):
# logits = self.logits(torch.cat([hidden_states[i], h], 1))
# logits = self.logits(nn.functional.relu(self.z_decoder(hidden_states[i], logits)))
logits = self.logits(
nn.functional.relu(
self.z_decoder(torch.cat([hidden_states[i], z], 1),
logits)))
# build the next z sample
if test:
q = OneHotCategorical(logits=logits)
z = q.sample()
else:
q = RelaxedOneHotCategorical(temperature=self.temp,
logits=logits)
z = q.rsample()
lse = log_sum_exp(logits, dim=1).view(-1, 1)
log_probs = logits - lse
# now, compute the log-likelihood of the data given this z-sample
# so emission is [batch_sz x z_dim], i.e. emission[i, j] is the log-probability of getting this
# data for element i given choice z
emission = F.embedding(input[i].repeat(n_particles), emit)
NLL = -log_sum_exp(emission + log_probs, 1)
nlls[i] = NLL.data
KL = (log_probs.exp() * (log_probs -
(prior_probs + 1e-16).log())).sum(1)
loss += (NLL + KL)
if i != seq_len - 1:
prior_probs = (T.unsqueeze(0) * z.unsqueeze(1)).sum(2)
# now, we calculate the final log-marginal estimator
return loss.sum(), nlls.sum(), (seq_len * batch_sz * n_particles), 0
class PPOTorchPolicy(TorchPolicy):
def __init__(self, observation_space, action_space, config):
super().__init__(observation_space, action_space, config)
self.device = torch.device('cpu')
# Get hyperparameters
self.alpha = config['alpha']
self.clip_ratio = config['clip_ratio']
self.gamma = config['gamma']
self.lam = config['lambda']
self.lr_pi = config['lr_pi']
self.lr_vf = config['lr_vf']
self.model_hidden_sizes = config['model_hidden_sizes']
self.num_skills = config['num_skills']
self.skill_input = config['skill_input']
self.target_kl = config['target_kl']
self.use_diayn = config['use_diayn']
self.use_env_rewards = config['use_env_rewards']
self.use_gae = config['use_gae']
# Initialize actor-critic model
self.skills = OneHotCategorical(torch.ones((1, self.num_skills)))
if self.skill_input is not None:
skill_vec = [0.] * (self.num_skills - 1)
skill_vec.insert(self.skill_input, 1.)
self.z = torch.as_tensor([skill_vec], dtype=torch.float32)
else:
self.z = None
self.model = SkilledA2C(observation_space,
action_space,
hidden_sizes=self.model_hidden_sizes,
skills=self.skills).to(self.device)
# Set up optimizers for policy and value function
self.pi_optimizer = Adam(self.model.pi.parameters(), self.lr_pi)
self.vf_optimizer = Adam(self.model.vf.parameters(), self.lr_vf)
self.disc_optimizer = Adam(self.model.discriminator.parameters(),
self.lr_vf)
def compute_loss_d(self, batch):
obs, z = batch[SampleBatch.CUR_OBS], batch[SKILLS]
logq_z = self.model.discriminator(obs)
return nn.functional.nll_loss(logq_z, z.argmax(dim=-1))
def compute_loss_pi(self, batch):
obs, act, z = batch[
SampleBatch.CUR_OBS], batch[ACTIVATIONS], batch[SKILLS]
adv, logp_old = batch[Postprocessing.ADVANTAGES], batch[
SampleBatch.ACTION_LOGP]
clip_ratio = self.clip_ratio
# Policy loss
oz = torch.cat([obs, z], dim=-1)
pi, logp = self.model.pi(oz, act)
ratio = torch.exp(logp - logp_old)
clip_adv = torch.clamp(ratio, 1 - clip_ratio, 1 + clip_ratio) * adv
loss_pi = -(torch.min(ratio * adv, clip_adv)).mean()
# Useful extra info
approx_kl = (logp_old - logp).mean().item()
ent = pi.entropy().mean().item()
clipped = ratio.gt(1 + clip_ratio) | ratio.lt(1 - clip_ratio)
clip_frac = torch.as_tensor(clipped, dtype=torch.float32).mean().item()
pi_info = dict(kl=approx_kl, ent=ent, cf=clip_frac)
return loss_pi, pi_info
def compute_loss_v(self, batch):
obs, z = batch[SampleBatch.NEXT_OBS], batch[SKILLS]
v_pred_old, v_targ = batch[SampleBatch.VF_PREDS], batch[
Postprocessing.VALUE_TARGETS]
oz = torch.cat([obs, z], dim=-1)
v_pred = self.model.vf(oz)
v_pred_clipped = v_pred_old + torch.clamp(
v_pred - v_pred_old, -self.clip_ratio, self.clip_ratio)
loss_clipped = (v_pred_clipped - v_targ).pow(2)
loss_unclipped = (v_pred - v_targ).pow(2)
return 0.5 * torch.max(loss_unclipped, loss_clipped).mean()
def _convert_activation_to_action(self, activation):
min_ = self.action_space.low
max_ = self.action_space.high
return tanh_to_action(activation, min_, max_)
def _normalize_obs(self, obs):
min_ = self.observation_space.low
max_ = self.observation_space.high
return normalize_obs(obs, min_, max_)
@override(Policy)
def compute_actions(self, obs, **kwargs):
# Sample a skill at the start of each episode
if self.z is None:
self.z = self.skills.sample()
o = self._normalize_obs(obs)
a, v, logp_a, logq_z = self.model.step(
torch.as_tensor(o, dtype=torch.float32), self.z)
actions = self._convert_activation_to_action(a)
extras = {
ACTIVATIONS: a,
SampleBatch.VF_PREDS: v,
SampleBatch.ACTION_LOGP: logp_a,
SKILLS: self.z.numpy(),
SKILL_LOGQ: logq_z
}
return actions, [], extras
@override(Policy)
def postprocess_trajectory(self,
batch,
other_agent_batches=None,
episode=None):
"""Adds the policy logits, VF preds, and advantages to the trajectory."""
completed = batch["dones"][-1]
if completed:
# Force end of episode reward
last_r = 0.0
# Reset skill at the end of each episode
self.z = None
else:
next_state = []
for i in range(self.num_state_tensors()):
next_state.append([batch["state_out_{}".format(i)][-1]])
obs = [batch[SampleBatch.NEXT_OBS][-1]]
o = self._normalize_obs(obs)
_, last_r, _, _ = self.model.step(
torch.as_tensor(o, dtype=torch.float32), self.z)
last_r = last_r.item()
# Compute DIAYN rewards
if self.use_diayn:
z = torch.as_tensor(batch[SKILLS], dtype=torch.float32)
logp_z = self.skills.log_prob(z).numpy()
logq_z = batch[SKILL_LOGQ][:, z.argmax(dim=-1)[0].item()]
entropy_reg = self.alpha * batch[SampleBatch.ACTION_LOGP]
diayn_rewards = logq_z - logp_z - entropy_reg
if self.use_env_rewards:
batch[SampleBatch.REWARDS] += diayn_rewards
else:
batch[SampleBatch.REWARDS] = diayn_rewards
batch = compute_advantages(batch,
last_r,
gamma=self.gamma,
lambda_=self.lam,
use_gae=self.use_gae)
return batch
@override(Policy)
def learn_on_batch(self, postprocessed_batch):
postprocessed_batch[SampleBatch.CUR_OBS] = self._normalize_obs(
postprocessed_batch[SampleBatch.CUR_OBS])
train_batch = self._lazy_tensor_dict(postprocessed_batch)
# Train policy with multiple steps of gradient descent
self.pi_optimizer.zero_grad()
loss_pi, pi_info = self.compute_loss_pi(train_batch)
# if pi_info['kl'] > 1.5 * self.target_kl:
# logger.info('Early stopping at step %d due to reaching max kl.' % i)
# return
loss_pi.backward()
self.pi_optimizer.step()
# Value function learning
self.vf_optimizer.zero_grad()
loss_v = self.compute_loss_v(train_batch)
loss_v.backward()
self.vf_optimizer.step()
# Discriminator learning
self.disc_optimizer.zero_grad()
loss_d = self.compute_loss_d(train_batch)
loss_d.backward()
self.disc_optimizer.step()
grad_info = dict(pi_loss=loss_pi.item(),
vf_loss=loss_v.item(),
d_loss=loss_d.item(),
**pi_info)
return {LEARNER_STATS_KEY: grad_info}
def sample(self,
obs,
prev_acts,
rnn_hidden_states,
available_actions=None,
sample_gumbel=False):
# TODO: review this method
act_logits, h_outs = self.forward(obs, prev_acts, rnn_hidden_states)
if self.multidiscrete:
sampled_actions = []
mean_action_logprobs = []
max_prob_actions = []
for act_logit in act_logits:
categorical = OneHotCategorical(logits=act_logit)
all_action_prob = categorical.probs
eps = (all_action_prob == 0.0) * 1e-6
all_action_logprob = torch.log(all_action_prob +
eps.float().detach())
mean_action_logprob = (all_action_logprob *
all_action_prob).sum(
dim=-1).unsqueeze(-1)
if sample_gumbel:
# get a differentiable sample of the action
sampled_action = gumbel_softmax(act_logit, hard=True)
else:
sampled_action = categorical.sample()
max_prob_action = onehot_from_logits(act_logit)
sampled_actions.append(sampled_action)
mean_action_logprobs.append(mean_action_logprob)
max_prob_actions.append(max_prob_action)
sampled_actions = torch.cat(sampled_actions, dim=-1)
mean_action_logprobs = torch.cat(mean_action_logprobs, dim=-1)
max_prob_actions = torch.cat(max_prob_actions, dim=-1)
return sampled_actions, mean_action_logprobs, max_prob_actions, h_outs
else:
categorical = OneHotCategorical(logits=act_logits)
all_action_probs = categorical.probs
eps = (all_action_probs == 0.0) * 1e-6
all_action_logprobs = torch.log(all_action_probs +
eps.float().detach())
mean_action_logprobs = (all_action_logprobs *
all_action_probs).sum(dim=-1).unsqueeze(-1)
if sample_gumbel:
# get a differentiable sample of the action
sampled_actions = gumbel_softmax(act_logits,
available_actions,
hard=True)
else:
if available_actions is not None:
if type(available_actions) == np.ndarray:
available_actions = torch.from_numpy(available_actions)
act_logits[available_actions == 0] = -1e10
sampled_actions = OneHotCategorical(
logits=act_logits).sample()
else:
sampled_actions = categorical.sample()
max_prob_actions = onehot_from_logits(act_logits,
available_actions)
return sampled_actions, mean_action_logprobs, max_prob_actions, h_outs
def sampled_filter(self, input, args, n_particles, emb, hidden_states):
seq_len, batch_sz = input.size()
T = F.log_softmax(self.T, 0) # NOTE: in log-space
pi = F.log_softmax(self.pi, 0) # NOTE: in log-space
emit = self.calc_emit()
hidden_states = hidden_states.repeat(1, n_particles, 1)
nlls = hidden_states.data.new(seq_len, batch_sz * n_particles)
loss = 0
accumulated_weights = -math.log(
n_particles) # will contain log w_{t - 1}
resamples = 0
# in log probability space
prior_logits = pi.unsqueeze(0).expand(batch_sz * n_particles,
self.z_dim)
for i in range(seq_len):
# the approximate posterior comes from the same thing as before
logits = self.logits(hidden_states[i])
if not self.training:
# this is crucial!!
p = OneHotCategorical(logits=prior_logits)
q = OneHotCategorical(logits=logits)
z = q.sample()
else:
p = RelaxedOneHotCategorical(temperature=self.temp_prior,
logits=prior_logits)
q = RelaxedOneHotCategorical(temperature=self.temp,
logits=logits)
z = q.rsample()
# now, compute the log-likelihood of the data given this z-sample
emission = F.embedding(input[i].repeat(n_particles), emit)
NLL = -(emission * z).sum(1)
# NLL = -self.decode(z, input[i].repeat(n_particles), (emit,)) # diff. w.r.t. z
nlls[i] = NLL.data
# compute the weight using `reweight` on page (4)
f_term = p.log_prob(z) # prior
r_term = q.log_prob(z) # proposal
alpha = -NLL + (f_term - r_term)
wa = accumulated_weights + alpha.view(n_particles, batch_sz)
Z = log_sum_exp(wa, dim=0) # line 7
loss += Z # line 8
accumulated_weights = wa - Z # F.log_softmax(wa, dim=0) # line 9
# sample ancestors, and reindex everything
if args.filter:
probs = accumulated_weights.data.exp()
probs += 0.01
probs = probs / probs.sum(0, keepdim=True)
effective_sample_size = 1. / probs.pow(2).sum(0)
# probs is [n_particles, batch_sz]
# ancestors [2 x 15] = [[0, 0, 0, ..., 0], [0, 1, 2, 3, ...]]
# offsets [2 x 15] = [[0, 0, 0, ..., 0], [1, 1, 1, 1, ...]]
# resample / RSAMP
if ((effective_sample_size / n_particles) < 0.3).sum() > 0:
resamples += 1
ancestors = torch.multinomial(probs.transpose(0, 1),
n_particles, True)
# now, reindex, which is the most important thing
offsets = n_particles * torch.arange(batch_sz).unsqueeze(
1).repeat(1, n_particles).long()
if ancestors.is_cuda:
offsets = offsets.cuda()
unrolled_idx = Variable(ancestors + offsets).view(-1)
z = torch.index_select(z, 0, unrolled_idx)
# reset accumulated_weights
accumulated_weights = -math.log(
n_particles) # will contain log w_{t - 1}
if i != seq_len - 1:
# now in log-probability space
prior_logits = log_sum_exp(T.unsqueeze(0) + z.unsqueeze(1), 2)
if self.training:
(-loss.sum() /
(seq_len * batch_sz * n_particles)).backward(retain_graph=True)
return -loss.sum(), nlls.sum(), seq_len * batch_sz * n_particles, 0
def forward(self, input, args, n_particles, test=False):
"""
evaluation is the IWAE-10 bound
"""
if test:
n_particles = 10
else:
n_particles = 1
pi = F.log_softmax(self.pi, 0)
# run the input and teacher-forcing inputs through the embedding layers here
seq_len, batch_sz = input.size()
emb = self.inp_embedding(input)
hidden = self.init_hidden(batch_sz, self.nhid, 2) # bidirectional
hidden_states, (_, _) = self.encoder(emb, hidden)
hidden_states = hidden_states.repeat(1, n_particles, 1)
# run the z-decoder at this point, evaluating the NLL at each step
h = (Variable(
hidden_states.data.new(batch_sz * n_particles,
self.hidden_size).zero_()),
Variable(
hidden_states.data.new(batch_sz * n_particles,
self.hidden_size).zero_()))
nlls = hidden_states.data.new(seq_len, batch_sz * n_particles)
loss = 0
# now a log-prob
prior_logits = pi.unsqueeze(0).expand(batch_sz * n_particles,
self.z_dim)
prior_h = (Variable(torch.zeros(batch_sz * n_particles, 50).cuda()),
Variable(torch.zeros(batch_sz * n_particles, 50).cuda()))
logits = self.init_hidden(batch_sz * n_particles,
self.z_dim,
squeeze=True)
feed = None
x_emb = self.lockdrop(emb, self.dropout_x)
for i in range(seq_len):
# build the next z sample - not differentiable! we don't train the inference network
logits = F.log_softmax(self.logits(hidden_states[i]), 1)
if test:
q = OneHotCategorical(logits=logits)
# p = OneHotCategorical(logits=prior_logits)
z = q.sample()
else:
q = RelaxedOneHotCategorical(temperature=self.temp,
logits=logits)
# p = RelaxedOneHotCategorical(temperature=self.temp_prior, logits=prior_logits)
z = q.rsample()
# this should be batch_sz x x_dim
scores = torch.mm(self.project(torch.cat([h[0], z], 1)),
self.emit.t())
NLL = nn.CrossEntropyLoss(reduce=False)(
scores, input[i].repeat(n_particles))
# KL = q.log_prob(z) - p.log_prob(z)
KL = (logits.exp() * (logits - prior_logits)).sum(1)
loss += (NLL + KL)
# else:
# loss += (NLL + args.anneal * KL)
nlls[i] = NLL.data
# set things up for next time
if i != seq_len - 1:
feed = torch.cat(
[emb[i].repeat(n_particles, 1),
self.z_emb(z)], 1)
prior_h = self.z_decoder(feed, prior_h)
prior_logits = F.log_softmax(self.project_z(prior_h[0]), 1)
h = self.hidden_rnn(x_emb[i].repeat(n_particles, 1),
h) # feed the next word into the RNN
if n_particles != 1:
loss = -log_sum_exp(-loss.view(n_particles, batch_sz),
0) + math.log(n_particles)
NLL = -log_sum_exp(
-nlls.view(seq_len, n_particles, batch_sz), 1) + math.log(
n_particles) # not quite accurate, but what can you do
else:
NLL = nlls
# now, we calculate the final log-marginal estimator
return loss.sum(), NLL.sum(), (seq_len * batch_sz), 0
