Remove unneeded casts to bool (#21983)

Remove cast to Bool

src/transformers/models/decision_transformer/modeling_decision_transformer.py

@@ -231,7 +231,7 @@ class DecisionTransformerGPT2Attention(nn.Module):
         if not self.is_cross_attention:
             # if only "normal" attention layer implements causal mask
             query_length, key_length = query.size(-2), key.size(-2)
-            causal_mask = self.bias[:, :, key_length - query_length : key_length, :key_length].bool()
+            causal_mask = self.bias[:, :, key_length - query_length : key_length, :key_length]
             mask_value = torch.finfo(attn_weights.dtype).min
             # Need to be a tensor, otherwise we get error: `RuntimeError: expected scalar type float but found double`.
             # Need to be on the same device, otherwise `RuntimeError: ..., x and y to be on the same device`

src/transformers/models/gpt2/modeling_gpt2.py

@@ -243,7 +243,7 @@ class GPT2Attention(nn.Module):
         if not self.is_cross_attention:
             # if only "normal" attention layer implements causal mask
             query_length, key_length = query.size(-2), key.size(-2)
-            causal_mask = self.bias[:, :, key_length - query_length : key_length, :key_length].bool()
+            causal_mask = self.bias[:, :, key_length - query_length : key_length, :key_length]
             mask_value = torch.finfo(attn_weights.dtype).min
             # Need to be a tensor, otherwise we get error: `RuntimeError: expected scalar type float but found double`.
             # Need to be on the same device, otherwise `RuntimeError: ..., x and y to be on the same device`

src/transformers/models/imagegpt/modeling_imagegpt.py

@@ -294,7 +294,7 @@ class ImageGPTAttention(nn.Module):
         if not self.is_cross_attention:
             # if only "normal" attention layer implements causal mask
             query_length, key_length = query.size(-2), key.size(-2)
-            causal_mask = self.bias[:, :, key_length - query_length : key_length, :key_length].bool()
+            causal_mask = self.bias[:, :, key_length - query_length : key_length, :key_length]
             mask_value = torch.finfo(attn_weights.dtype).min
             # Need to be a tensor, otherwise we get error: `RuntimeError: expected scalar type float but found double`.
             # Need to be on the same device, otherwise `RuntimeError: ..., x and y to be on the same device`