refactor : replace iteration with consumed_samples

- Replace context.iteration with consumed_samples (global sample count) - Add optimizer_step property derived from consumed_samples - Checkpoint meta.json stores consumed_samples, drops iteration - CLI --start_batch renamed to --start_samples (per-rank samples) - Checkpoint dir naming: epoch_X_step_Y instead of epoch_X_iter_Y - Metric log entries use step and consumed_samples fields - Backward compat removed (old iteration checkpoints unsupported)
2026-06-30 18:42:42 +08:00
parent 44579ea6dc
commit aabb0d83e9
10 changed files with 72 additions and 48 deletions
@@ -74,7 +74,9 @@ class Trainer:
                        context.loss = loss.item()
                        stand_loss = loss / executor.grad_accum_steps
                        executor.backward(stand_loss)
-                        context.iteration += 1
+                        context.consumed_samples += (
+                            context.config.batch_per_device * context.world_size
+                        )
                        self._call_callbacks("on_batch_end", context)

                        if executor.sync_gradients: