gpt-bigcode: avoid `zero_` to support Core ML (#24755)

gpt-bigcode: avoid `zeros_` to support Core ML.

In-place `zeros_` is not supported by the Core ML conversion process.
This PR replaces it with `zeros_like` so conversion can proceed.

The change only affects a workaround for a PyTorch bug on the `cpu`
device.

src/transformers/models/gpt_bigcode/modeling_gpt_bigcode.py

@@ -164,7 +164,7 @@ class GPTBigCodeAttention(nn.Module):
             # This is needed because of a bug in pytorch https://github.com/pytorch/pytorch/issues/80588.
             # The bug was fixed in https://github.com/pytorch/pytorch/pull/96086,
             # but the fix has not been released as of pytorch version 2.0.0.
-            attn_weights.zero_()
+            attn_weights = torch.zeros_like(attn_weights)
             beta = 1
         else:
             beta = 0