internal

PiperOrigin-RevId: 468794403

seqio/beam_utils.py

@@ -359,11 +359,8 @@ class GetStats(beam.PTransform):
   prefixed by the identifiers.
   """
 
-  def __init__(self,
-               output_features: Mapping[str, seqio.Feature],
-               task_ids: Optional[Mapping[str, int]] = None):
+  def __init__(self, output_features: Mapping[str, seqio.Feature]):
     self._output_features = output_features
-    self._task_ids = task_ids or {}
 
   def expand(self, pcoll):
     example_counts = (
@@ -399,15 +396,6 @@ def _merge_dicts(dicts):
         merged_dict.update(d)
       return merged_dict
 
-    stats = [example_counts, total_tokens, max_tokens, char_length]
-    if self._task_ids:
-      task_ids_dict = {"task_ids": self._task_ids}
-      task_ids = (
-          pcoll
-          | "sample_for_task_ids" >> beam.combiners.Sample.FixedSizeGlobally(1)
-          | "create_task_ids" >> beam.Map(lambda _: task_ids_dict))
-      stats.append(task_ids)
-
-    return (stats
+    return ([example_counts, total_tokens, max_tokens, char_length]
             | "flatten_counts" >> beam.Flatten()
             | "merge_stats" >> beam.CombineGlobally(_merge_dicts))

seqio/beam_utils_test.py

@@ -222,49 +222,6 @@ def test_get_stats_tokenized_dataset(self):
               "targets_chars": 12,
           }]))
 
-  def test_get_stats_task_ids(self):
-    # These examples are assumed to be decoded by
-    # `seqio.test_utils.sentencepiece_vocab()`.
-    input_examples = [{
-        # Decoded as "ea", i.e., length 2 string
-        "inputs": np.array([4, 5]),
-        # Decoded as "ea test", i.e., length 7 string
-        "targets": np.array([4, 5, 10]),
-    }, {
-        # Decoded as "e", i.e., length 1 string
-        "inputs": np.array([4]),
-        # Decoded as "asoil", i.e., length 5 string. "1" is an EOS id.
-        "targets": np.array([5, 6, 7, 8, 9, 1])
-    }]
-
-    output_features = seqio.test_utils.FakeTaskTest.DEFAULT_OUTPUT_FEATURES
-    with TestPipeline() as p:
-      pcoll = (
-          p
-          | beam.Create(input_examples)
-          | beam_utils.GetStats(
-              output_features=output_features,
-              task_ids={
-                  "task_name_1": 1,
-                  "task_name_2": 2
-              }))
-
-      util.assert_that(
-          pcoll,
-          util.equal_to([{
-              "inputs_tokens": 3,  # 4 and 3 from the first and second exmaples.
-              "targets_tokens": 8,
-              "inputs_max_tokens": 2,
-              "targets_max_tokens": 5,
-              "examples": 2,
-              "inputs_chars": 3,
-              "targets_chars": 12,
-              "task_ids": {
-                  "task_name_1": 1,
-                  "task_name_2": 2
-              }
-          }]))
-
   def test_count_characters_tokenized_dataset(self):
     # These examples are assumed to be decoded by
     # `seqio.test_utils.sentencepiece_vocab()`.

seqio/test_data/cached_task_with_provenance/stats.train.json

@@ -5,8 +5,5 @@
   "inputs_tokens": 36,
   "targets_chars": 29,
   "targets_max_tokens": 6,
-  "targets_tokens": 18,
-  "task_ids": {
-    "task_name": 1
-  }
+  "targets_tokens": 18
 }

seqio/test_data/cached_task_with_provenance/stats.validation.json

@@ -5,8 +5,5 @@
   "inputs_tokens": 23,
   "targets_chars": 37,
   "targets_max_tokens": 21,
-  "targets_tokens": 36,
-  "task_ids": {
-    "task_name": 1
-  }
+  "targets_tokens": 36
 }