Fix MPS backend 'index out of range' error (lm-sys#2737)

fastchat/model/model_adapter.py

@@ -212,8 +212,16 @@ def load_model(
                 kwargs["max_memory"] = {i: max_gpu_memory for i in range(num_gpus)}
     elif device == "mps":
         kwargs = {"torch_dtype": torch.float16}
-        # Avoid bugs in mps backend by not using in-place operations.
-        replace_llama_attn_with_non_inplace_operations()
+        import transformers
+
+        version = tuple(int(v) for v in transformers.__version__.split("."))
+        if version < (4, 35, 0):
+            # NOTE: Recent transformers library seems to fix the mps issue, also
+            # it has made some changes causing compatibility issues with our
+            # original patch. So we only apply the patch for older versions.
+
+            # Avoid bugs in mps backend by not using in-place operations.
+            replace_llama_attn_with_non_inplace_operations()
     elif device == "xpu":
         kwargs = {"torch_dtype": torch.bfloat16}
         # Try to load ipex, while it looks unused, it links into torch for xpu support