self.batch_size = batch_size
self.device = device
+ self.height = height
+ self.width = width
states, actions, rewards = escape.generate_episodes(
nb_train_samples + nb_test_samples, height, width, T
)
- seq = escape.episodes2seq(states, actions, rewards)
- self.train_input = seq[:nb_train_samples]
- self.test_input = seq[nb_train_samples:]
+ seq = escape.episodes2seq(states, actions, rewards, lookahead_delta=5)
+ self.train_input = seq[:nb_train_samples].to(self.device)
+ self.test_input = seq[nb_train_samples:].to(self.device)
self.nb_codes = max(self.train_input.max(), self.test_input.max()) + 1
def produce_results(
self, n_epoch, model, result_dir, logger, deterministic_synthesis, nmax=1000
):
- pass
+ result = self.test_input[:100].clone()
+ ar_mask = (
+ torch.arange(result.size(1), device=result.device)
+ > self.height * self.width + 2
+ ).long()[None, :]
+ ar_mask = ar_mask.expand_as(result)
+ result *= 1 - ar_mask # paraaaaanoiaaaaaaa
+
+ masked_inplace_autoregression(
+ model,
+ self.batch_size,
+ result,
+ ar_mask,
+ deterministic_synthesis,
+ device=self.device,
+ )
+
+ s, a, r, lr = escape.seq2episodes(
+ result, self.height, self.width, lookahead=True
+ )
+ str = escape.episodes2str(
+ s, a, r, lookahead_rewards=lr, unicode=True, ansi_colors=True
+ )
+
+ filename = os.path.join(result_dir, f"test_seq_{n_epoch:04d}.txt")
+ with open(filename, "w") as f:
+ f.write(str)
+ logger(f"wrote {filename}")
######################################################################