Fix resize embedding with Deepspeed (#32192)

fix resize when deepspeed

src/transformers/modeling_utils.py

@@ -2131,13 +2131,23 @@ def _get_resized_embeddings(
         # Replace weights in old_embeddings and return to maintain the same embedding type.
         # This ensures correct functionality when a Custom Embedding class is passed as input.
         # The input and output embedding types remain consistent. (c.f. https://github.com/huggingface/transformers/pull/31979)
-        old_embeddings.weight.data = new_embeddings.weight.data
-        old_embeddings.num_embeddings = new_embeddings.weight.data.shape[0]
+        if is_deepspeed_zero3_enabled() and not is_quantized:
+            import deepspeed
+
+            params = [old_embeddings.weight, new_embeddings.weight]
+            with deepspeed.zero.GatheredParameters(params, modifier_rank=0):
+                old_embeddings.weight.data = new_embeddings.weight.data
+                old_embeddings.num_embeddings = new_embeddings.weight.data.shape[0]
 
-        # If the new number of tokens is smaller than the original `padding_idx`, the `padding_idx`
-        # will be set to `None` in the resized embeddings.
-        if old_embeddings.padding_idx is not None and (new_num_tokens - 1) < old_embeddings.padding_idx:
-            old_embeddings.padding_idx = None
+                # If the new number of tokens is smaller than the original `padding_idx`, the `padding_idx`
+                # will be set to `None` in the resized embeddings.
+                if old_embeddings.padding_idx is not None and (new_num_tokens - 1) < old_embeddings.padding_idx:
+                    old_embeddings.padding_idx = None
+        else:
+            old_embeddings.weight.data = new_embeddings.weight.data
+            old_embeddings.num_embeddings = new_embeddings.weight.data.shape[0]
+            if old_embeddings.padding_idx is not None and (new_num_tokens - 1) < old_embeddings.padding_idx:
+                old_embeddings.padding_idx = None
 
         return old_embeddings