cd_binary_binary: added gibbs sampling

src/rbm/entity.py

@@ -102,7 +102,6 @@ class Entity:
 		state = self.state.v_to_h(v)
 		return state
 
-
 	def v_given_h(self, h: Mat) -> Mat:
 		state = self.state.h_to_v(h)
 		return state

src/rbm/train.py

@@ -100,8 +100,10 @@ def cd_binary_binary(entity: Entity, data_pos: Mat, params: TrainingParams):
 	data_neg = prob(entity.v_given_h(h_probs_pos))
 	h_probs_neg = prob(entity.h_given_v(data_neg))
 
-	# Gibbs	sampling with training params
+	# Gibbs	sampling
-	# ToDo
+	for _ in range(params.num_gibbs_samples-1):
+		data_neg = prob(entity.v_given_h(h_probs_neg))
+		h_probs_neg = prob(entity.h_given_v(data_neg))
 
 	# Update weights (negative phase)
 	dw -= np.dot(np.transpose(data_neg), h_probs_neg)

src/tests/test_learn_encoded_labels.py

@@ -10,7 +10,7 @@ from rbm.train import TrainingParams
 class LabelLearner(Model):
 	def __init__(self, name: str, work_dir: str = '.'):
 		super().__init__(name, work_dir)
-		self.unit1 = Entity((16, 16), EntityParams())
+		self.unit1 = Entity((16, 8), EntityParams())
 
 	def forward(self, x: Mat):
 		x = self.unit1.forward(x)
@@ -45,7 +45,7 @@ if __name__ == "__main__":
 	model.load()
 
 	# Train
-	model.train(encoded, TrainingParams(learning_rate=0.01, momentum=0.9, do_rao_blackwell=True, num_epochs=10000))
+	model.train(encoded, TrainingParams(learning_rate=0.01, momentum=0.9, do_rao_blackwell=True, num_epochs=1000))
 
 	# save state
 	model.save()
@@ -57,7 +57,7 @@ if __name__ == "__main__":
 		axes[index].imshow(img)
 		axes[index].axis('off')
 		axes[index].set_title(f'{test_labels[index]}')
-
+		print(inp)
 	plt.show()
 
 	print("Test: [passed]")