From 3c34382265510ef0e9b02695bcc3552f83f7d891 Mon Sep 17 00:00:00 2001
From: Cemberk <Cem.bozkus@amd.com>
Date: Tue, 3 Dec 2024 09:41:49 -0600
Subject: [PATCH] fix conflict in trainer.py (#75)

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
---
 src/transformers/trainer.py | 68 ++-----------------------------------
 1 file changed, 3 insertions(+), 65 deletions(-)

diff --git a/src/transformers/trainer.py b/src/transformers/trainer.py
index c4d777266bf58f..1ba7ed01236903 100755
--- a/src/transformers/trainer.py
+++ b/src/transformers/trainer.py
@@ -2483,73 +2483,11 @@ def _inner_training_loop(
                     if not do_sync_step:
                         self.accelerator.gradient_state._set_sync_gradients(False)
                     else:
-                        self.state.num_input_tokens_seen += (
-                            torch.sum(
-                                self.accelerator.gather(
-                                    torch.tensor(
-                                        inputs[main_input_name].numel(), device=self.args.device, dtype=torch.int64
-                                    )
-                                )
-                            )
-                            .cpu()
-                            .item()
-                        )
-                if rng_to_sync:
-                    self._load_rng_state(resume_from_checkpoint)
-                    rng_to_sync = False
-
-                if (self.state.global_step == 10):
-                    start_train_stable_time = time.time()
-
-                # Skip past any already trained steps if resuming training
-                if steps_trained_in_current_epoch > 0:
-                    steps_trained_in_current_epoch -= 1
-                    if steps_trained_progress_bar is not None:
-                        steps_trained_progress_bar.update(1)
-                    if steps_trained_in_current_epoch == 0:
-                        self._load_rng_state(resume_from_checkpoint)
-                    continue
-                elif steps_trained_progress_bar is not None:
-                    steps_trained_progress_bar.close()
-                    steps_trained_progress_bar = None
-
-                if step % args.gradient_accumulation_steps == 0:
-                    self.control = self.callback_handler.on_step_begin(args, self.state, self.control)
-
-                with self.accelerator.accumulate(model):
-                    tr_loss_step = self.training_step(model, inputs)
-
-                if (
-                    args.logging_nan_inf_filter
-                    and not is_torch_xla_available()
-                    and (torch.isnan(tr_loss_step) or torch.isinf(tr_loss_step))
-                ):
-                    # if loss is nan or inf simply add the average of previous logged losses
-                    tr_loss += tr_loss / (1 + self.state.global_step - self._globalstep_last_logged)
-                else:
-                    if tr_loss.device != tr_loss_step.device:
-                        raise ValueError(
-                            f"Calculated loss must be on the original device: {tr_loss.device} but device in use is {tr_loss_step.device}"
-                        )
-                    tr_loss += tr_loss_step
-
-                self.current_flos += float(self.floating_point_ops(inputs))
-
-                is_last_step_and_steps_less_than_grad_acc = (
-                    steps_in_epoch <= args.gradient_accumulation_steps and (step + 1) == steps_in_epoch
-                )
-
-                if (
-                    total_batched_samples % args.gradient_accumulation_steps == 0
-                    or
-                    # last step in epoch but step is always smaller than gradient_accumulation_steps
-                    is_last_step_and_steps_less_than_grad_acc
-                ):
-                    # the `or` condition of `is_last_step_and_steps_less_than_grad_acc` is not covered
-                    # in accelerate. So, explicitly enable sync gradients to True in that case.
-                    if is_last_step_and_steps_less_than_grad_acc:
                         self.accelerator.gradient_state._set_sync_gradients(True)
 
+                    if (self.state.global_step == 10):
+                        start_train_stable_time = time.time()
+
                     if self.args.include_num_input_tokens_seen:
                         main_input_name = getattr(self.model, "main_input_name", "input_ids")
                         if main_input_name not in inputs: