update seeds

pytorch · Jan 22, 2025 · ee02a46 · ee02a46
1 parent 1060f3c
commit ee02a46
Showing 1 changed file with 40 additions and 17 deletions.
diff --git a/test/nodes/test_multi_node_weighted_sampler.py b/test/nodes/test_multi_node_weighted_sampler.py
@@ -24,7 +24,6 @@ def setUp(self) -> None:
         self._num_datasets = 4
         self._weights_fn = lambda i: 0.1 * (i + 1)
         self._num_epochs = 3
-        self._seed = 42
         self.datasets = {
             f"ds{i}": IterableWrapper(DummyIterableDataset(self._num_samples, f"ds{i}"))
             for i in range(self._num_datasets)
@@ -39,7 +38,7 @@ def test_torchdata_nodes_imports(self) -> None:
             self.fail("MultiNodeWeightedSampler or StopCriteria failed to import")
 
     def _setup_multi_node_weighted_sampler(
-        self, num_samples, num_datasets, weights_fn, stop_criteria, seed
+        self, num_samples, num_datasets, weights_fn, stop_criteria, seed=0
     ) -> Prefetcher:
 
         datasets = {f"ds{i}": IterableWrapper(DummyIterableDataset(num_samples, f"ds{i}")) for i in range(num_datasets)}
@@ -54,7 +53,8 @@ def test_multi_node_weighted_sampler_weight_sampler_keys_mismatch(self) -> None:
             "keys of source_nodes and weights must be the same",
         ):
             MultiNodeWeightedSampler(
-                self.datasets, {f"dummy{i}": self._weights_fn(i) for i in range(self._num_datasets)}, seed=self._seed
+                self.datasets,
+                {f"dummy{i}": self._weights_fn(i) for i in range(self._num_datasets)},
             )
 
     def test_multi_node_weighted_batch_sampler_invalid_weights_tensor_shape(
@@ -63,7 +63,8 @@ def test_multi_node_weighted_batch_sampler_invalid_weights_tensor_shape(
         """Test validation logic for MultiNodeWeightedSampler if the shape of the weights tensor is invalid"""
         with self.assertRaisesRegex(ValueError, " weights must be a 1d sequence, non-negative, and non-zero"):
             MultiNodeWeightedSampler(
-                self.datasets, weights={f"ds{i}": [[1.0]] for i in range(self._num_datasets)}, seed=self._seed
+                self.datasets,
+                weights={f"ds{i}": [[1.0]] for i in range(self._num_datasets)},
             )
 
     def test_multi_node_weighted_batch_sampler_negative_weights(
@@ -72,7 +73,8 @@ def test_multi_node_weighted_batch_sampler_negative_weights(
         """Test validation logic for MultiNodeWeightedSampler if the value of the weights tensor is invalid"""
         with self.assertRaisesRegex(ValueError, " weights must be a 1d sequence, non-negative, and non-zero"):
             MultiNodeWeightedSampler(
-                self.datasets, weights={f"ds{i}": -1 for i in range(self._num_datasets)}, seed=self._seed
+                self.datasets,
+                weights={f"ds{i}": -1 for i in range(self._num_datasets)},
             )
 
     def test_multi_node_weighted_batch_sampler_zero_weights(
@@ -81,10 +83,11 @@ def test_multi_node_weighted_batch_sampler_zero_weights(
         """Test validation logic for MultiNodeWeightedSampler if the value of the weights tensor is invalid"""
         with self.assertRaisesRegex(ValueError, " weights must be a 1d sequence, non-negative, and non-zero"):
             MultiNodeWeightedSampler(
-                self.datasets, weights={f"ds{i}": 10 * i for i in range(self._num_datasets)}, seed=self._seed
+                self.datasets,
+                weights={f"ds{i}": 10 * i for i in range(self._num_datasets)},
             )
 
-    @parameterized.expand([0, 1, 42])
+    @parameterized.expand(range(10))
     def test_multi_node_weighted_sampler_first_exhausted(self, seed) -> None:
         """Test MultiNodeWeightedSampler with stop criteria FIRST_DATASET_EXHAUSTED"""
         mixer = self._setup_multi_node_weighted_sampler(
@@ -108,7 +111,7 @@ def test_multi_node_weighted_sampler_first_exhausted(self, seed) -> None:
             self.assertGreaterEqual(dataset_counts_in_results.count(self._num_samples), 1)
             mixer.reset()
 
-    @parameterized.expand([0, 1, 42])
+    @parameterized.expand(range(10))
     def test_multi_node_weighted_sampler_all_dataset_exhausted(self, seed) -> None:
         """Test MultiNodeWeightedSampler with stop criteria ALL_DATASETS_EXHAUSTED"""
         mixer = self._setup_multi_node_weighted_sampler(
@@ -135,7 +138,7 @@ def test_multi_node_weighted_sampler_all_dataset_exhausted(self, seed) -> None:
             self.assertEqual(sorted(set(datasets_in_results)), ["ds0", "ds1", "ds2", "ds3"])
             mixer.reset()
 
-    @parameterized.expand([0, 1, 42])
+    @parameterized.expand(range(10))
     def test_multi_node_weighted_sampler_cycle_until_all_exhausted(self, seed) -> None:
         """Test MultiNodeWeightedSampler with stop criteria CYCLE_UNTIL_ALL_DATASETS_EXHAUSTED"""
         mixer = self._setup_multi_node_weighted_sampler(
@@ -154,7 +157,7 @@ def test_multi_node_weighted_sampler_cycle_until_all_exhausted(self, seed) -> No
             self.assertEqual(sorted(datasets_in_results), ["ds0", "ds1", "ds2", "ds3"])
             mixer.reset()
 
-    @parameterized.expand([0, 1, 42])
+    @parameterized.expand(range(10))
     def test_multi_node_weighted_sampler_cycle_forever(self, seed) -> None:
         """Test MultiNodeWeightedSampler with stop criteria CYCLE_FOREVER"""
         mixer = MultiNodeWeightedSampler(
@@ -173,7 +176,12 @@ def test_multi_node_weighted_sampler_cycle_forever(self, seed) -> None:
     @parameterized.expand([(1, 8), (8, 32)])
     def test_multi_node_weighted_batch_sampler_set_rank_world_size(self, rank, world_size):
         """Test MultiNodeWeightedSampler with different rank and world size"""
-        mixer = MultiNodeWeightedSampler(self.datasets, self.weights, rank=rank, world_size=world_size, seed=self._seed)
+        mixer = MultiNodeWeightedSampler(
+            self.datasets,
+            self.weights,
+            rank=rank,
+            world_size=world_size,
+        )
         self.assertEqual(mixer.rank, rank)
         self.assertEqual(mixer.world_size, world_size)
 
@@ -183,7 +191,10 @@ def test_multi_node_weighted_batch_sampler_results_for_ranks(self):
         global_results = []
         for rank in range(world_size):
             mixer = MultiNodeWeightedSampler(
-                self.datasets, self.weights, rank=rank, world_size=world_size, seed=self._seed
+                self.datasets,
+                self.weights,
+                rank=rank,
+                world_size=world_size,
             )
             results = list(mixer)
             global_results.append(results)
@@ -198,7 +209,10 @@ def test_multi_node_weighted_batch_sampler_results_for_multiple_epochs(self):
         """Test MultiNodeWeightedSampler with different results in each epoch"""
 
         # Check for the mixer node only
-        mixer = MultiNodeWeightedSampler(self.datasets, self.weights, seed=self._seed)
+        mixer = MultiNodeWeightedSampler(
+            self.datasets,
+            self.weights,
+        )
 
         overall_results = []
         for _ in range(self._num_epochs):
@@ -219,7 +233,6 @@ def test_multi_node_weighted_batch_sampler_results_for_multiple_epochs(self):
             self._num_datasets,
             self._weights_fn,
             stop_criteria=StopCriteria.CYCLE_UNTIL_ALL_DATASETS_EXHAUSTED,
-            seed=self._seed,
         )
 
         overall_results = []
@@ -247,7 +260,11 @@ def test_multi_node_weighted_batch_sampler_results_for_multiple_epochs(self):
     )
     def test_save_load_state_mixer_over_multiple_epochs(self, midpoint: int, stop_criteria: str):
         """Test MultiNodeWeightedSampler with saving and loading of state across multiple epochs"""
-        node = MultiNodeWeightedSampler(self.datasets, self.weights, stop_criteria, seed=self._seed)
+        node = MultiNodeWeightedSampler(
+            self.datasets,
+            self.weights,
+            stop_criteria,
+        )
         run_test_save_load_state(self, node, midpoint)
 
     @parameterized.expand(
@@ -262,7 +279,10 @@ def test_save_load_state_mixer_over_multiple_epochs(self, midpoint: int, stop_cr
     )
     def test_save_load_state_mixer_over_multiple_epochs_with_prefetcher(self, midpoint: int, stop_criteria: str):
         node = self._setup_multi_node_weighted_sampler(
-            self._num_samples, self._num_datasets, self._weights_fn, stop_criteria=stop_criteria, seed=self._seed
+            self._num_samples,
+            self._num_datasets,
+            self._weights_fn,
+            stop_criteria=stop_criteria,
         )
         run_test_save_load_state(self, node, midpoint)
 
@@ -282,6 +302,9 @@ def test_multi_node_weighted_large_sample_size_with_prefetcher(self, midpoint, s
         num_datasets = 5
 
         mixer = self._setup_multi_node_weighted_sampler(
-            num_samples, num_datasets, self._weights_fn, stop_criteria, seed=self._seed
+            num_samples,
+            num_datasets,
+            self._weights_fn,
+            stop_criteria,
         )
         run_test_save_load_state(self, mixer, midpoint)