fix graph breaks

recipes/lora_finetune_single_device.py

@@ -664,6 +664,7 @@ def train(self) -> None:
             )
 
         # Initialize tokens count and running loss (for grad accumulation)
+        start = time.perf_counter()
         t0 = time.perf_counter()
         running_loss = 0
         num_tokens = 0
@@ -730,6 +731,7 @@ def train(self) -> None:
                         # Log per-step metrics
                         if self.global_step % self._log_every_n_steps == 0:
                             time_per_step = time.perf_counter() - t0
+                            print(time_per_step)
                             log_dict = {
                                 "loss": loss_to_log,
                                 "lr": self._optimizer.param_groups[0]["lr"],
@@ -773,13 +775,17 @@ def train(self) -> None:
                 self.epochs_run += 1
                 start_save_checkpoint = time.perf_counter()
                 log.info("Starting checkpoint save...")
-                self.save_checkpoint(epoch=curr_epoch)
+                # self.save_checkpoint(epoch=curr_epoch)
                 log.info(
                     "Checkpoint saved in {:.2f} seconds.".format(
                         time.perf_counter() - start_save_checkpoint
                     )
                 )
 
+        end = time.perf_counter()
+        time_total = end - start
+        print(f"{time_total=}")
+
     def cleanup(self) -> None:
         self._metric_logger.close()
 

torchtune/modules/loss/ce_chunked_output_loss.py

@@ -78,6 +78,12 @@ def forward(self, logits: List[torch.Tensor], labels: torch.Tensor) -> torch.Ten
         # compute one chunk at a time
         total_loss = 0.0
         for logits_chunk, labels_chunk in zip(logits, labels):
+
+            # avoid graph breaks when seq_len is not constant in the batch
+            torch._dynamo.mark_dynamic(logits_chunk, 0)
+            torch._dynamo.mark_dynamic(labels_chunk, 0)
+
+            # CE
             total_loss += self.compute_cross_entropy(logits_chunk, labels_chunk)
 
         return total_loss / total_elements

torchtune/modules/transformer.py

@@ -628,14 +628,23 @@ def forward(
 
         # shape: [b, s, d]
         h = self.tok_embeddings(tokens)
+        h.requires_grad = True  # avoid graph breaks when using LoRA
 
         hidden = []
         for i, layer in enumerate(self.layers):
             if i in self.output_hidden_states:
                 hidden.append(h)
 
+            # avoid graph breaks when seq_len is not constant in the batch
+            torch._dynamo.mark_dynamic(h, 1)
+            if mask is not None:
+                torch._dynamo.mark_dynamic(mask, 1)
+            if encoder_mask is not None:
+                torch._dynamo.mark_dynamic(encoder_mask, 1)
+            if input_pos is not None:
+                torch._dynamo.mark_dynamic(input_pos, 1)
+
             # shape: [b, s, d]
-            torch._dynamo.mark_dynamic(h, 1)  # avoid graph breaks
             h = layer(
                 h,
                 mask=mask,