fix fsdp auto wrap policy (#1302)

* fix fsdp policy

* fix fsdp

* revert

* refactor to be inline with Accelerate

src/peft/utils/other.py

@@ -360,6 +360,20 @@ def fsdp_auto_wrap_policy(model):
 
     from ..tuners import PrefixEncoder, PromptEmbedding, PromptEncoder
 
+    default_transformer_cls_names_to_wrap = (
+        ",".join(model._no_split_modules) if getattr(model, "_no_split_modules", None) is not None else ""
+    )
+    transformer_cls_names_to_wrap = os.environ.get(
+        "FSDP_TRANSFORMER_CLS_TO_WRAP", default_transformer_cls_names_to_wrap
+    ).split(",")
+    transformer_cls_to_wrap = {PrefixEncoder, PromptEncoder, PromptEmbedding}
+    for layer_class in transformer_cls_names_to_wrap:
+        transformer_cls = FullyShardedDataParallelPlugin.get_module_class_from_name(model, layer_class)
+        if transformer_cls is None:
+            raise Exception("Could not find the transformer layer class to wrap in the model.")
+        else:
+            transformer_cls_to_wrap.add(transformer_cls)
+
     def lambda_policy_fn(module):
         if (
             len(list(module.named_children())) == 0
@@ -372,14 +386,7 @@ def lambda_policy_fn(module):
     lambda_policy = functools.partial(lambda_auto_wrap_policy, lambda_fn=lambda_policy_fn)
     transformer_wrap_policy = functools.partial(
         transformer_auto_wrap_policy,
-        transformer_layer_cls=(
-            PrefixEncoder,
-            PromptEncoder,
-            PromptEmbedding,
-            FullyShardedDataParallelPlugin.get_module_class_from_name(
-                model, os.environ.get("FSDP_TRANSFORMER_CLS_TO_WRAP", "")
-            ),
-        ),
+        transformer_layer_cls=transformer_cls_to_wrap,
     )
 
     auto_wrap_policy = functools.partial(_or_policy, policies=[lambda_policy, transformer_wrap_policy])