Resolve DeepSpeed cannot resume training with PeftModel (#28746)

* fix: resolve deepspeed resume peft model issues

* chore: update something

* chore: update model instance pass into is peft model checks

* chore: remove hard code value to tests

* fix: format code

src/transformers/integrations/deepspeed.py

@@ -143,14 +143,25 @@ def trainer_config_process(self, args, auto_find_batch_size=False):
             "per_device_train_batch_size",
             not auto_find_batch_size,
         )
-        self.fill_match("gradient_accumulation_steps", args.gradient_accumulation_steps, "gradient_accumulation_steps")
         self.fill_match(
-            "train_batch_size", train_batch_size, "train_batch_size (calculated)", not auto_find_batch_size
+            "gradient_accumulation_steps",
+            args.gradient_accumulation_steps,
+            "gradient_accumulation_steps",
+        )
+        self.fill_match(
+            "train_batch_size",
+            train_batch_size,
+            "train_batch_size (calculated)",
+            not auto_find_batch_size,
         )
         self.fill_match("gradient_clipping", args.max_grad_norm, "max_grad_norm")
 
         self.fill_match("optimizer.params.lr", args.learning_rate, "learning_rate")
-        self.fill_match("optimizer.params.betas", [args.adam_beta1, args.adam_beta2], "adam_beta1+adam_beta2")
+        self.fill_match(
+            "optimizer.params.betas",
+            [args.adam_beta1, args.adam_beta2],
+            "adam_beta1+adam_beta2",
+        )
         self.fill_match("optimizer.params.eps", args.adam_epsilon, "adam_epsilon")
         self.fill_match("optimizer.params.weight_decay", args.weight_decay, "weight_decay")
 
@@ -225,12 +236,26 @@ def trainer_config_finalize(self, args, model, num_training_steps):
             self.fill_only("zero_optimization.reduce_bucket_size", hidden_size * hidden_size)
             if self.is_zero3():
                 # automatically assign the optimal config values based on model config
-                self.fill_only("zero_optimization.stage3_prefetch_bucket_size", 0.9 * hidden_size * hidden_size)
-                self.fill_only("zero_optimization.stage3_param_persistence_threshold", 10 * hidden_size)
+                self.fill_only(
+                    "zero_optimization.stage3_prefetch_bucket_size",
+                    0.9 * hidden_size * hidden_size,
+                )
+                self.fill_only(
+                    "zero_optimization.stage3_param_persistence_threshold",
+                    10 * hidden_size,
+                )
 
         # scheduler
-        self.fill_match("scheduler.params.total_num_steps", num_training_steps, "num_training_steps (calculated)")
-        self.fill_match("scheduler.params.warmup_num_steps", args.get_warmup_steps(num_training_steps), "warmup_steps")
+        self.fill_match(
+            "scheduler.params.total_num_steps",
+            num_training_steps,
+            "num_training_steps (calculated)",
+        )
+        self.fill_match(
+            "scheduler.params.warmup_num_steps",
+            args.get_warmup_steps(num_training_steps),
+            "warmup_steps",
+        )
 
         if len(self.mismatches) > 0:
             mismatches = "\n".join(self.mismatches)
@@ -387,7 +412,7 @@ def deepspeed_init(trainer, num_training_steps, inference=False):
     return optimizer, lr_scheduler
 
 
-def deepspeed_load_checkpoint(deepspeed_engine, checkpoint_path):
+def deepspeed_load_checkpoint(deepspeed_engine, checkpoint_path, load_module_strict=True):
     # it's possible that the user is trying to resume from model_path, which doesn't necessarily
     # contain a deepspeed checkpoint. e.g. examples just check if the dir exists and assume it's
     # a resume from a checkpoint and not just a local pretrained weight. So we check here if the
@@ -400,7 +425,10 @@ def deepspeed_load_checkpoint(deepspeed_engine, checkpoint_path):
         logger.info(f"Attempting to resume from {checkpoint_path}")
         # this magically updates self.optimizer and self.lr_scheduler
         load_path, _ = deepspeed_engine.load_checkpoint(
-            checkpoint_path, load_optimizer_states=True, load_lr_scheduler_states=True
+            checkpoint_path,
+            load_module_strict=load_module_strict,
+            load_optimizer_states=True,
+            load_lr_scheduler_states=True,
         )
         if load_path is None:
             raise ValueError(f"[deepspeed] failed to resume from checkpoint {checkpoint_path}")

src/transformers/trainer.py

@@ -1727,7 +1727,9 @@ def _inner_training_loop(
         # ckpt loading
         if resume_from_checkpoint is not None:
             if self.is_deepspeed_enabled:
-                deepspeed_load_checkpoint(self.model_wrapped, resume_from_checkpoint)
+                deepspeed_load_checkpoint(
+                    self.model_wrapped, resume_from_checkpoint, load_module_strict=not _is_peft_model(self.model)
+                )
             elif is_sagemaker_mp_enabled() or self.is_fsdp_enabled:
                 self._load_from_checkpoint(resume_from_checkpoint, self.model_wrapped)
 
@@ -2193,7 +2195,11 @@ def _load_best_model(self):
 
         model = self.model_wrapped if is_sagemaker_mp_enabled() else self.model
         if self.is_deepspeed_enabled:
-            deepspeed_load_checkpoint(self.model_wrapped, self.state.best_model_checkpoint)
+            deepspeed_load_checkpoint(
+                self.model_wrapped,
+                self.state.best_model_checkpoint,
+                load_module_strict=not _is_peft_model(self.model),
+            )
         elif self.is_fsdp_enabled:
             load_result = load_fsdp_model(
                 self.accelerator.state.fsdp_plugin,