Removing unnecessary `device=device` in modeling_llama.py (#24696)

* Update modeling_llama.py

Removing unnecessary `device=device`

* fix in all occurrences of _make_causal_mask

src/transformers/models/autoformer/modeling_autoformer.py

@@ -365,7 +365,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/bart/modeling_bart.py

@@ -93,7 +93,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/bigbird_pegasus/modeling_bigbird_pegasus.py

@@ -85,7 +85,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/biogpt/modeling_biogpt.py

@@ -61,7 +61,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/blenderbot/modeling_blenderbot.py

@@ -83,7 +83,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/blenderbot_small/modeling_blenderbot_small.py

@@ -79,7 +79,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/clip/modeling_clip.py

@@ -682,7 +682,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/clipseg/modeling_clipseg.py

@@ -692,7 +692,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/groupvit/modeling_groupvit.py

@@ -1075,7 +1075,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/informer/modeling_informer.py

@@ -238,7 +238,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/led/modeling_led.py

@@ -79,7 +79,7 @@ def _make_causal_mask(input_ids_shape: torch.Size, dtype: torch.dtype, past_key_
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min))
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min)
     mask_cond = torch.arange(mask.size(-1))
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/llama/modeling_llama.py

@@ -46,7 +46,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/m2m_100/modeling_m2m_100.py

@@ -79,7 +79,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/marian/modeling_marian.py

@@ -81,7 +81,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/mbart/modeling_mbart.py

@@ -86,7 +86,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/musicgen/modeling_musicgen.py

@@ -86,7 +86,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/mvp/modeling_mvp.py

@@ -97,7 +97,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/nllb_moe/modeling_nllb_moe.py

@@ -84,7 +84,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/open_llama/modeling_open_llama.py

@@ -55,7 +55,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)

src/transformers/models/opt/modeling_opt.py

@@ -71,7 +71,7 @@ def _make_causal_mask(
     Make causal mask used for bi-directional self-attention.
     """
     bsz, tgt_len = input_ids_shape
-    mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min, device=device), device=device)
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
     mask_cond = torch.arange(mask.size(-1), device=device)
     mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
     mask = mask.to(dtype)