diff --git a/tripy/pyproject.toml b/tripy/pyproject.toml
index 6e6a9c91d..43f6bf6ee 100644
--- a/tripy/pyproject.toml
+++ b/tripy/pyproject.toml
@@ -8,8 +8,8 @@ requires-python = ">= 3.9"
 license = {text = "Apache 2.0"}
 dependencies = [
   "tensorrt~=10.0",
-  "mlir-tensorrt-compiler==0.1.34+cuda12.trt102",
-  "mlir-tensorrt-runtime==0.1.34+cuda12.trt102",
+  "mlir-tensorrt-compiler==0.1.36+cuda12.trt102",
+  "mlir-tensorrt-runtime==0.1.36+cuda12.trt102",
   "colored==2.2.3",
 ]
 
diff --git a/tripy/tests/integration/test_iota.py b/tripy/tests/integration/test_iota.py
index 39df35787..2df779da2 100644
--- a/tripy/tests/integration/test_iota.py
+++ b/tripy/tests/integration/test_iota.py
@@ -91,9 +91,7 @@ def test_negative_no_casting(self, dtype):
         a = tp.ones((2, 2))
         out = Iota.build([frontend_utils.tensor_from_shape_like(a.shape)], dim=0, output_rank=2, dtype=dtype)
 
-        exception_str = "error: 'tensorrt.linspace' op result #0 must be 0D/1D/2D/3D/4D/5D/6D/7D/8D tensor of 32-bit float or 32-bit signless integer values"
-        if dtype == tp.bool:
-            exception_str = "InternalError: failed to run compilation"
+        exception_str = "InternalError: failed to run compilation"
         with helper.raises(
             tp.TripyException,
             match=exception_str,
diff --git a/tripy/tests/integration/test_quantize.py b/tripy/tests/integration/test_quantize.py
index b50293869..826db83bf 100644
--- a/tripy/tests/integration/test_quantize.py
+++ b/tripy/tests/integration/test_quantize.py
@@ -23,6 +23,7 @@
 import tripy as tp
 from tests.helper import raises, TORCH_DTYPES
 from tests.conftest import skip_if_older_than_sm80, skip_if_older_than_sm89
+import cupy as cp
 
 
 class TestQuantize:
@@ -118,4 +119,4 @@ def test_non_constant_scale(self):
         scale = tp.ones((4,))
         quantized = tp.quantize(input, scale, tp.int8, dim=0)
 
-        assert bool(tp.all(quantized == tp.ones((4, 4), dtype=tp.int8)))
+        assert bool(cp.all(cp.from_dlpack(quantized) == cp.ones((4, 4), dtype=cp.int8)))
diff --git a/tripy/tripy/frontend/ops/tensor_initializers.py b/tripy/tripy/frontend/ops/tensor_initializers.py
index da5b1e221..faa395f7a 100644
--- a/tripy/tripy/frontend/ops/tensor_initializers.py
+++ b/tripy/tripy/frontend/ops/tensor_initializers.py
@@ -165,7 +165,7 @@ def zeros_like(input: "tripy.Tensor", dtype: Optional[datatype.dtype] = None) ->
 @constraints.dtypes(
     constraints={"tensor": "T1", constraints.RETURN_VALUE: "T1"},
     variables={
-        "T1": ["float32", "float16", "bfloat16", "int32", "bool"],
+        "T1": ["float32", "float16", "bfloat16", "int32", "int64", "bool"],
     },
 )
 def tril(tensor: "tripy.Tensor", diagonal: int = 0) -> "tripy.Tensor":
@@ -223,7 +223,7 @@ def tril(tensor: "tripy.Tensor", diagonal: int = 0) -> "tripy.Tensor":
 @constraints.dtypes(
     constraints={"tensor": "T1", constraints.RETURN_VALUE: "T1"},
     variables={
-        "T1": ["float32", "float16", "bfloat16", "int32", "bool"],
+        "T1": ["float32", "float16", "bfloat16", "int32", "int64", "bool"],
     },
 )
 def triu(tensor: "tripy.Tensor", diagonal: int = 0) -> "tripy.Tensor":
@@ -281,7 +281,7 @@ def triu(tensor: "tripy.Tensor", diagonal: int = 0) -> "tripy.Tensor":
 @constraints.dtypes(
     constraints={"dtype": "T1", constraints.RETURN_VALUE: "T1"},
     variables={
-        "T1": ["float32", "float16", "bfloat16", "int8", "int32", "bool"],
+        "T1": ["float32", "float16", "bfloat16", "int8", "int32", "int64", "bool"],
     },
 )
 def arange(
@@ -346,7 +346,7 @@ def arange(
 @constraints.dtypes(
     constraints={"dtype": "T1", constraints.RETURN_VALUE: "T1"},
     variables={
-        "T1": ["float32", "float16", "bfloat16", "int8", "int32", "bool"],
+        "T1": ["float32", "float16", "bfloat16", "int8", "int32", "int64", "bool"],
     },
 )
 def arange(
diff --git a/tripy/tripy/frontend/trace/ops/flip.py b/tripy/tripy/frontend/trace/ops/flip.py
index 15e12b2fa..ab9820174 100644
--- a/tripy/tripy/frontend/trace/ops/flip.py
+++ b/tripy/tripy/frontend/trace/ops/flip.py
@@ -40,7 +40,7 @@ def to_flat_ir(self, inputs, outputs):
 @constraints.dtypes(
     constraints={"input": "T1", constraints.RETURN_VALUE: "T1"},
     variables={
-        "T1": ["float32", "float16", "bfloat16", "int32", "bool"],
+        "T1": ["float32", "float16", "bfloat16", "int32", "int64", "bool"],
     },
 )
 def flip(input: "tripy.Tensor", dims: Optional[Union[int, Sequence[int]]] = None) -> "tripy.Tensor":
diff --git a/tripy/tripy/frontend/trace/ops/gather.py b/tripy/tripy/frontend/trace/ops/gather.py
index 0df478585..d18294d91 100644
--- a/tripy/tripy/frontend/trace/ops/gather.py
+++ b/tripy/tripy/frontend/trace/ops/gather.py
@@ -91,7 +91,7 @@ def to_flat_ir(self, inputs, outputs):
 @constraints.dtypes(
     constraints={"input": "T1", "index": "T2", constraints.RETURN_VALUE: "T1"},
     variables={
-        "T1": ["float32", "float16", "bfloat16", "int8", "int32", "bool"],
+        "T1": ["float32", "float16", "bfloat16", "int4", "int8", "int32", "int64", "bool"],
         "T2": ["int32"],
     },
 )
diff --git a/tripy/tripy/frontend/trace/ops/iota.py b/tripy/tripy/frontend/trace/ops/iota.py
index a252d756c..98933f92a 100644
--- a/tripy/tripy/frontend/trace/ops/iota.py
+++ b/tripy/tripy/frontend/trace/ops/iota.py
@@ -69,7 +69,7 @@ def iota_impl(shape: "tripy.Tensor", dim: int, dtype: datatype.dtype, output_ran
 @constraints.dtypes(
     constraints={"dtype": "T1", constraints.RETURN_VALUE: "T1"},
     variables={
-        "T1": ["float32", "float16", "bfloat16", "float8", "int4", "int8", "int32", "bool"],
+        "T1": ["float32", "float16", "bfloat16", "float8", "int4", "int8", "int32", "int64", "bool"],
     },
 )
 def iota(shape: ShapeLike, dim: int = 0, dtype: datatype.dtype = datatype.float32) -> "tripy.Tensor":
@@ -101,7 +101,7 @@ def iota(shape: ShapeLike, dim: int = 0, dtype: datatype.dtype = datatype.float3
     constraints={"input": "T1", "dtype": "T2", constraints.RETURN_VALUE: "T2"},
     variables={
         "T1": ["float32", "float16", "bfloat16", "float8", "int4", "int8", "int32", "int64", "bool"],
-        "T2": ["float32", "float16", "bfloat16", "float8", "int4", "int8", "int32", "bool"],
+        "T2": ["float32", "float16", "bfloat16", "float8", "int4", "int8", "int32", "int64", "bool"],
     },
 )
 def iota_like(input: "tripy.Tensor", dim: int = 0, dtype: Optional[datatype.dtype] = None) -> "tripy.Tensor":
diff --git a/tripy/tripy/frontend/trace/ops/reduce.py b/tripy/tripy/frontend/trace/ops/reduce.py
index e12a8129a..cba001f78 100644
--- a/tripy/tripy/frontend/trace/ops/reduce.py
+++ b/tripy/tripy/frontend/trace/ops/reduce.py
@@ -135,7 +135,7 @@ def _reduce_impl(input: "tripy.Tensor", kind: Reduce.Kind, dim: Union[int, Seque
 @export.public_api(document_under="operations/functions")
 @constraints.dtypes(
     constraints={"input": "T1", constraints.RETURN_VALUE: "T1"},
-    variables={"T1": ["float32", "int32", "float16", "bfloat16"]},
+    variables={"T1": ["float32", "int32", "int64", "float16", "bfloat16"]},
 )
 def sum(
     input: "tripy.Tensor", dim: Optional[Union[int, Sequence[int]]] = None, keepdim: bool = False
@@ -232,7 +232,7 @@ def any(
 @export.public_api(document_under="operations/functions")
 @constraints.dtypes(
     constraints={"input": "T1", constraints.RETURN_VALUE: "T1"},
-    variables={"T1": ["float32", "int32", "float16", "bfloat16"]},
+    variables={"T1": ["float32", "int32", "int64", "float16", "bfloat16"]},
 )
 def max(
     input: "tripy.Tensor", dim: Optional[Union[int, Sequence[int]]] = None, keepdim: bool = False
@@ -265,7 +265,7 @@ def max(
 @export.public_api(document_under="operations/functions")
 @constraints.dtypes(
     constraints={"input": "T1", constraints.RETURN_VALUE: "T1"},
-    variables={"T1": ["float32", "int32", "float16", "bfloat16"]},
+    variables={"T1": ["float32", "int32", "int64", "float16", "bfloat16"]},
 )
 def prod(
     input: "tripy.Tensor", dim: Optional[Union[int, Sequence[int]]] = None, keepdim: bool = False
@@ -313,7 +313,7 @@ def mean_impl(tensor: "tripy.Tensor", dim: Union[int, Sequence] = None, keepdim:
 @export.public_api(document_under="operations/functions")
 @constraints.dtypes(
     constraints={"input": "T1", constraints.RETURN_VALUE: "T1"},
-    variables={"T1": ["float32", "int32", "float16", "bfloat16"]},
+    variables={"T1": ["float32", "int32", "int64", "float16", "bfloat16"]},
 )
 def mean(
     input: "tripy.Tensor", dim: Optional[Union[int, Sequence[int]]] = None, keepdim: bool = False
@@ -413,7 +413,7 @@ def _arg_min_max_impl(tensor: "tripy.Tensor", kind: ArgMinMax.Kind, dim: Optiona
 @export.public_api(document_under="operations/functions")
 @constraints.dtypes(
     constraints={"input": "T1", constraints.RETURN_VALUE: "T2"},
-    variables={"T1": ["float32", "float16", "bfloat16", "int32", "bool", "int8"], "T2": ["int32"]},
+    variables={"T1": ["float32", "float16", "bfloat16", "int32"], "T2": ["int32"]},
 )
 def argmax(input: "tripy.Tensor", dim: Optional[int] = None, keepdim: bool = False) -> "tripy.Tensor":
     """
@@ -445,7 +445,7 @@ def argmax(input: "tripy.Tensor", dim: Optional[int] = None, keepdim: bool = Fal
 @export.public_api(document_under="operations/functions")
 @constraints.dtypes(
     constraints={"input": "T1", constraints.RETURN_VALUE: "T2"},
-    variables={"T1": ["float32", "float16", "bfloat16", "int32", "bool", "int8"], "T2": ["int32"]},
+    variables={"T1": ["float32", "float16", "bfloat16", "int32"], "T2": ["int32"]},
 )
 def argmin(input: "tripy.Tensor", dim: Optional[int] = None, keepdim: bool = False) -> "tripy.Tensor":
     """
diff --git a/tripy/tripy/frontend/trace/ops/unary_elementwise.py b/tripy/tripy/frontend/trace/ops/unary_elementwise.py
index 08e630f0c..bb2ad3d00 100644
--- a/tripy/tripy/frontend/trace/ops/unary_elementwise.py
+++ b/tripy/tripy/frontend/trace/ops/unary_elementwise.py
@@ -252,7 +252,7 @@ def log(input: "tripy.Tensor") -> "tripy.Tensor":
 @export.public_api(document_under="operations/functions")
 @constraints.dtypes(
     constraints={"input": "T1", constraints.RETURN_VALUE: "T1"},
-    variables={"T1": ["float32", "float16", "bfloat16", "int8", "int32"]},
+    variables={"T1": ["float32", "float16", "bfloat16", "int8", "int32", "int64"]},
 )
 def abs(input: "tripy.Tensor") -> "tripy.Tensor":
     r"""