revert batching mode (#2479)

* revert batching mode
OpenNMT · Sep 26, 2023 · d61467e · d61467e
1 parent e6a4412
commit d61467e
Show file tree

Hide file tree

Showing 2 changed files with 19 additions and 17 deletions.
diff --git a/onmt/inputters/dynamic_iterator.py b/onmt/inputters/dynamic_iterator.py
@@ -279,27 +279,30 @@ def batch_size_fn(nbsents, maxlen):
             else:
                 raise ValueError(f"Invalid argument batch_type={batch_type}")
 
-        minibatch, maxlen, size_so_far, seen = [], 0, 0, []
+        def max_src_tgt(ex):
+            """return the max tokens btw src and tgt in the sequence."""
+            if ex["tgt"]:
+                return max(len(ex["src"]["src_ids"]), len(ex["tgt"]["tgt_ids"]))
+            return len(ex["src"]["src_ids"])
+
+        minibatch, maxlen, size_so_far, seen = [], 0, 0, set()
         for ex in data:
-            if (ex["src"]["src"] not in seen) or (self.task != CorpusTask.TRAIN):
-                seen.append(ex["src"]["src"])
+            src = ex["src"]["src"]
+            if src not in seen or (self.task != CorpusTask.TRAIN):
+                seen.add(src)
                 minibatch.append(ex)
                 nbsents = len(minibatch)
-                maxlen = max(text_sort_key(ex), maxlen)
+                maxlen = max(max_src_tgt(ex), maxlen)
                 size_so_far = batch_size_fn(nbsents, maxlen)
                 if size_so_far >= batch_size:
-                    overflowed = 0
-                    if size_so_far > batch_size:
-                        overflowed += 1
+                    overflowed = 1 if size_so_far > batch_size else 0
                     if batch_size_multiple > 1:
-                        overflowed += (
-                            len(minibatch) - overflowed
-                        ) % batch_size_multiple
+                        overflowed += (nbsents - overflowed) % batch_size_multiple
                     if overflowed == 0:
                         yield minibatch
-                        minibatch, maxlen, size_so_far, seen = [], 0, 0, []
+                        minibatch, maxlen, size_so_far, seen = [], 0, 0, set()
                     else:
-                        if overflowed == len(minibatch):
+                        if overflowed == nbsents:
                             logger.warning(
                                 "The batch will be filled until we reach"
                                 " %d, its size may exceed %d tokens"
@@ -308,10 +311,9 @@ def batch_size_fn(nbsents, maxlen):
                         else:
                             yield minibatch[:-overflowed]
                             minibatch = minibatch[-overflowed:]
-                            maxlen, size_so_far, seen = 0, 0, []
-                            for i, ex in enumerate(minibatch):
-                                maxlen = max(text_sort_key(ex), maxlen)
-                                size_so_far = batch_size_fn(i + 1, maxlen)
+                            maxlen = max([max_src_tgt(ex) for ex in minibatch])
+                            size_so_far = batch_size_fn(len(minibatch), maxlen)
+                            seen = set()
 
         if minibatch:
             yield minibatch

diff --git a/onmt/inputters/text_utils.py b/onmt/inputters/text_utils.py
@@ -59,7 +59,7 @@ def append_features_to_text(text, features):
 def text_sort_key(ex):
     """Sort using the number of tokens in the sequence."""
     if ex["tgt"]:
-        return max(len(ex["src"]["src_ids"]), len(ex["tgt"]["tgt_ids"]))
+        return len(ex["src"]["src_ids"]), len(ex["tgt"]["tgt_ids"])
     return len(ex["src"]["src_ids"])