Fix mistake in total steps computation for GLUE notebook (#142)

* fix to total_steps formula
* Update .meta.yml
* drop cpu

Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>

Author: Valahaar

lightning_examples/text-transformers/.meta.yml

@@ -1,9 +1,9 @@
 title: Finetune Transformers Models with PyTorch Lightning
 author: PL team
 created: 2021-01-31
-updated: 2021-12-03
+updated: 2022-02-08
 license: CC BY-SA
-build: 2
+build: 0
 tags:
   - Text
 description: |
@@ -17,5 +17,4 @@ requirements:
   - scikit-learn
   - torchtext>=0.9
 accelerator:
-  - CPU
   - GPU

lightning_examples/text-transformers/text-transformers.py

@@ -224,8 +224,8 @@ def setup(self, stage=None) -> None:
 
         # Calculate total steps
         tb_size = self.hparams.train_batch_size * max(1, self.trainer.gpus)
-        ab_size = self.trainer.accumulate_grad_batches * float(self.trainer.max_epochs)
-        self.total_steps = (len(train_loader.dataset) // tb_size) // ab_size
+        ab_size = tb_size * self.trainer.accumulate_grad_batches
+        self.total_steps = int((len(train_loader.dataset) / ab_size) * float(self.trainer.max_epochs))
 
     def configure_optimizers(self):
         """Prepare optimizer and schedule (linear warmup and decay)"""