Merge pull request #520 from mlcommons/juhan/criteo_size_fix

Update number of data points for Criteo
mlcommons · Sep 25, 2023 · ae3587d · ae3587d
2 parents 92b2d1d + 86ad0af
commit ae3587d
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 3 deletions.
diff --git a/algorithmic_efficiency/workloads/criteo1tb/criteo1tb_pytorch/workload.py b/algorithmic_efficiency/workloads/criteo1tb/criteo1tb_pytorch/workload.py
@@ -233,7 +233,7 @@ def _eval_batch(self,
     summed_loss = self.loss_fn(
         label_batch=batch['targets'], logits_batch=logits,
         mask_batch=weights)['summed']
-    return summed_loss
+    return summed_loss.to(dtype=torch.float64)
 
 
 class Criteo1TbDlrmSmallTestWorkload(Criteo1TbDlrmSmallWorkload):

diff --git a/algorithmic_efficiency/workloads/criteo1tb/workload.py b/algorithmic_efficiency/workloads/criteo1tb/workload.py
@@ -63,11 +63,11 @@ def num_eval_train_examples(self) -> int:
 
   @property
   def num_validation_examples(self) -> int:
-    return 89_000_000
+    return 83_274_637
 
   @property
   def num_test_examples(self) -> int:
-    return 89_274_637
+    return 95_000_000
 
   @property
   def train_mean(self):