[Generation] remove leftover code from end-to-end compilation

src/transformers/models/bamba/modeling_bamba.py

@@ -43,7 +43,6 @@
 from ...utils import (
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
-    is_torchdynamo_compiling,
     logging,
     replace_return_docstrings,
 )
@@ -1590,7 +1589,7 @@ def prepare_inputs_for_generation(
         if not empty_past_kv:
             if (
                 inputs_embeds is not None  # Exception 1
-                or (is_torchdynamo_compiling() or cache_position[-1] >= input_ids.shape[1])  # Exception 3
+                or cache_position[-1] >= input_ids.shape[1]  # Exception 3
             ):
                 input_ids = input_ids[:, -cache_position.shape[0] :]
             elif input_ids.shape[1] != cache_position.shape[0]:  # Default case (the "else", a no op, is Exception 2)

src/transformers/models/bamba/modular_bamba.py

@@ -51,7 +51,6 @@
 from ...utils import (
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
-    is_torchdynamo_compiling,
     logging,
     replace_return_docstrings,
 )
@@ -1274,7 +1273,7 @@ def prepare_inputs_for_generation(
         if not empty_past_kv:
             if (
                 inputs_embeds is not None  # Exception 1
-                or (is_torchdynamo_compiling() or cache_position[-1] >= input_ids.shape[1])  # Exception 3
+                or cache_position[-1] >= input_ids.shape[1]  # Exception 3
             ):
                 input_ids = input_ids[:, -cache_position.shape[0] :]
             elif input_ids.shape[1] != cache_position.shape[0]:  # Default case (the "else", a no op, is Exception 2)

src/transformers/models/bloom/modeling_bloom.py

@@ -38,7 +38,6 @@
 from ...modeling_utils import PreTrainedModel
 from ...utils import (
     is_torch_flex_attn_available,
-    is_torchdynamo_compiling,
     logging,
 )
 from .configuration_bloom import BloomConfig
@@ -919,7 +918,7 @@ def prepare_inputs_for_generation(
                 inputs_embeds = inputs_embeds[:, -cache_position.shape[0] :]
             elif (
                 inputs_embeds is not None  # Exception 1
-                or (is_torchdynamo_compiling() or cache_position[-1] >= input_ids.shape[1])  # Exception 3
+                or cache_position[-1] >= input_ids.shape[1]  # Exception 3
             ):
                 input_ids = input_ids[:, -cache_position.shape[0] :]
             elif input_ids.shape[1] != cache_position.shape[0]:  # Default case (the "else", a no op, is Exception 2)

src/transformers/models/cohere2/modeling_cohere2.py

@@ -36,7 +36,6 @@
     LossKwargs,
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
-    is_torchdynamo_compiling,
     logging,
     replace_return_docstrings,
 )
@@ -931,7 +930,7 @@ def prepare_inputs_for_generation(
         if past_key_values is not None:
             if (
                 inputs_embeds is not None  # Exception 1
-                or (is_torchdynamo_compiling() or cache_position[-1] >= input_ids.shape[1])  # Exception 3
+                or cache_position[-1] >= input_ids.shape[1]  # Exception 3
             ):
                 input_ids = input_ids[:, -cache_position.shape[0] :]
             elif input_ids.shape[1] != cache_position.shape[0]:  # Default case (the "else", a no op, is Exception 2)

src/transformers/models/cohere2/modular_cohere2.py

@@ -29,7 +29,6 @@
 from ...modeling_utils import ALL_ATTENTION_FUNCTIONS
 from ...processing_utils import Unpack
 from ...utils import (
-    is_torchdynamo_compiling,
     logging,
 )
 from ..cohere.modeling_cohere import (
@@ -603,7 +602,7 @@ def prepare_inputs_for_generation(
         if past_key_values is not None:
             if (
                 inputs_embeds is not None  # Exception 1
-                or (is_torchdynamo_compiling() or cache_position[-1] >= input_ids.shape[1])  # Exception 3
+                or cache_position[-1] >= input_ids.shape[1]  # Exception 3
             ):
                 input_ids = input_ids[:, -cache_position.shape[0] :]
             elif input_ids.shape[1] != cache_position.shape[0]:  # Default case (the "else", a no op, is Exception 2)

src/transformers/models/jamba/modeling_jamba.py

@@ -42,7 +42,6 @@
 from ...utils import (
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
-    is_torchdynamo_compiling,
     logging,
     replace_return_docstrings,
 )
@@ -1571,7 +1570,7 @@ def prepare_inputs_for_generation(
         if not empty_past_kv:
             if (
                 inputs_embeds is not None  # Exception 1
-                or (is_torchdynamo_compiling() or cache_position[-1] >= input_ids.shape[1])  # Exception 3
+                or cache_position[-1] >= input_ids.shape[1]  # Exception 3
             ):
                 input_ids = input_ids[:, -cache_position.shape[0] :]
             elif input_ids.shape[1] != cache_position.shape[0]:  # Default case (the "else", a no op, is Exception 2)

src/transformers/models/mllama/modeling_mllama.py

@@ -33,7 +33,6 @@
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
     is_torch_flex_attn_available,
-    is_torchdynamo_compiling,
     logging,
     replace_return_docstrings,
 )
@@ -2194,7 +2193,7 @@ def prepare_inputs_for_generation(
         if past_key_values is not None:
             if (
                 inputs_embeds is not None  # Exception 1
-                or (is_torchdynamo_compiling() or cache_position[-1] >= input_ids.shape[1])  # Exception 3
+                or cache_position[-1] >= input_ids.shape[1]  # Exception 3
             ):
                 input_ids = input_ids[:, -cache_position.shape[0] :]
             elif input_ids.shape[1] != cache_position.shape[0]:  # Default case (the "else", a no op, is Exception 2)

src/transformers/models/moshi/modeling_moshi.py

@@ -2507,7 +2507,7 @@ def prepare_inputs_for_generation(
         if past_key_values is not None:
             if (
                 inputs_embeds is not None  # Exception 1
-                or (is_torchdynamo_compiling() or cache_position[-1] >= input_ids.shape[1])  # Exception 3
+                or cache_position[-1] >= input_ids.shape[1]  # Exception 3
             ):
                 input_ids = input_ids[:, -cache_position.shape[0] :]
             elif input_ids.shape[1] != cache_position.shape[0]:  # Default case (the "else", a no op, is Exception 2)

src/transformers/models/zamba/modeling_zamba.py

@@ -45,7 +45,6 @@
 from ...utils import (
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
-    is_torchdynamo_compiling,
     logging,
     replace_return_docstrings,
 )
@@ -1325,7 +1324,7 @@ def prepare_inputs_for_generation(
             #              (we can't check exception 3 while compiling)
             if (
                 inputs_embeds is not None  # Exception 1
-                or (is_torchdynamo_compiling() or cache_position[-1] >= input_ids.shape[1])  # Exception 3
+                or cache_position[-1] >= input_ids.shape[1]  # Exception 3
             ):
                 input_ids = input_ids[:, -cache_position.shape[0] :]
             elif input_ids.shape[1] != cache_position.shape[0]:  # Default case (the "else", a no op, is Exception 2)

src/transformers/models/zamba2/modeling_zamba2.py

@@ -40,7 +40,6 @@
 from ...utils import (
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
-    is_torchdynamo_compiling,
     logging,
     replace_return_docstrings,
 )
@@ -1762,7 +1761,7 @@ def prepare_inputs_for_generation(
             #              (we can't check exception 3 while compiling)
             if (
                 inputs_embeds is not None  # Exception 1
-                or (is_torchdynamo_compiling() or cache_position[-1] >= input_ids.shape[1])  # Exception 3
+                or cache_position[-1] >= input_ids.shape[1]  # Exception 3
             ):
                 input_ids = input_ids[:, -cache_position.shape[0] :]
             elif input_ids.shape[1] != cache_position.shape[0]:  # Default case (the "else", a no op, is Exception 2)