Update.

[picoclvr.git] / main.py

diff --git a/main.py b/main.py
index c52881b..e1f619c 100755 (executable)
--- a/main.py
+++ b/main.py
@@ -1028,10 +1028,16 @@ class TaskExpr(Task):
         self.device = device
 
         train_sequences = expr.generate_sequences(
-            nb_train_samples, nb_variables=nb_variables, length=2*sequence_length, randomize_length=True,
+            nb_train_samples,
+            nb_variables=nb_variables,
+            length=sequence_length,
+            # length=2 * sequence_length,
+            # randomize_length=True,
         )
         test_sequences = expr.generate_sequences(
-            nb_test_samples, nb_variables=nb_variables, length=sequence_length,
+            nb_test_samples,
+            nb_variables=nb_variables,
+            length=sequence_length,
         )
         self.char2id = dict(
             [
@@ -1084,8 +1090,8 @@ class TaskExpr(Task):
             input.split(self.batch_size), dynamic_ncols=True, desc=desc
         ):
             if split == "train":
-                last=(batch!=self.filler).max(0).values.nonzero().max()+1
-                batch=batch[:,:last]
+                last = (batch != self.filler).max(0).values.nonzero().max() + 1
+                batch = batch[:, :last]
             yield batch
 
     def vocabulary_size(self):
@@ -1110,6 +1116,24 @@ class TaskExpr(Task):
                 nb_total = input.size(0)
                 nb_correct = (input == result).long().min(1).values.sum()
 
+                values_input = expr.extract_results([self.seq2str(s) for s in input])
+                max_input = max([max(x.values()) for x in values_input])
+                values_result = expr.extract_results([self.seq2str(s) for s in result])
+                max_result = max(
+                    [-1 if len(x) == 0 else max(x.values()) for x in values_result]
+                )
+
+                nb_missing, nb_predicted = torch.zeros(max_input + 1), torch.zeros(
+                    max_input + 1, max_result + 1
+                )
+                for i, r in zip(values_input, values_result):
+                    for n, vi in i.items():
+                        vr = r.get(n)
+                        if vr is None or vr < 0:
+                            nb_missing[vi] += 1
+                        else:
+                            nb_predicted[vi, vr] += 1
+
                 return nb_total, nb_correct
 
             test_nb_total, test_nb_correct = compute_nb_correct(self.test_input[:1000])
@@ -1131,7 +1155,7 @@ class TaskExpr(Task):
             )
             correct = (1 - ar_mask) * self.space + ar_mask * input
             for n in range(result.size(0)):
-                comment="GOOD" if (result[n]-input[n]).abs().max()==0 else ""
+                comment = "GOOD" if (result[n] - input[n]).abs().max() == 0 else ""
                 log_string(f"test_after  {self.seq2str(result[n])} {comment}")
                 log_string(f"correct     {self.seq2str(correct[n])}")
             ##############################################################