clip_norm=5.0)
buffer = ReplayBuffer(buffer_size=replay)
with tf.Session() as sess:
sess.run(model.tf_init())
losses = []
for epoch in range(1, NUM_EPOCHS + 1):
batch_losses = 0.0
for bTrainX, bTrainY in batch_gen.load_train():
if buffer.size < buffer.max_size:
buffer.add(state=bTrainX, action=bTrainY)
continue
else:
buffer.add(state=bTrainX, action=bTrainY)
state, reward, action = buffer.get_batch(bsz=BSZ)
loss = model.optimize(sess, state, action)
losses.append(loss)
print("Epoch {} Average Train Loss: {}, validating...".format(
epoch, np.mean(losses)))
losses = []
allocation_wts = []
price_change_vec = []
for bEvalX, bEvalYFat in batch_gen.load_test():
bEvalY = bEvalYFat[:, -1, :]
pred_allocations = model.predict_allocation(sess, bEvalX)
assert bEvalY.shape == pred_allocations.shape
price_change_vec.append(bEvalY)
allocation_wts.append(pred_allocations)
asset_list=ASSETS,
randomize_train=randomize_train,
overlapping_train=overlapping_train)
model = CNNModel(num_hid=NUM_HID,
bptt=BPTT,
num_assets=len(asset_list),
lr=LR,
clip_norm=5.0)
with tf.Session() as sess:
sess.run(model.tf_init())
losses = []
for epoch in range(1, NUM_EPOCHS + 1):
for bTrainX, bTrainY in batch_gen.load_train():
loss = model.optimize(sess, bTrainX, bTrainY)
losses.append(loss)
print("Epoch {} Average Train Loss: {}, validating...".format(
epoch,
sum(losses) / len(losses)))
losses = []
allocation_wts = []
price_change_vec = []
for bEvalX, bEvalYFat in batch_gen.load_test():
bEvalY = bEvalYFat[:, -1, :]
pred_allocations = model.predict_allocation(sess, bEvalX)
assert bEvalY.shape == pred_allocations.shape
price_change_vec.append(bEvalY)
allocation_wts.append(pred_allocations)
true_change_vec = np.concatenate(price_change_vec)