Build (aarch64)

build/torch26-cxx11-cu126-aarch64-linux/quantization_eetq/__init__.py

@@ -0,0 +1,3 @@
+from .custom_ops import w8_a16_gemm, w8_a16_gemm_, preprocess_weights, quant_weights
+
+__all__ = ["w8_a16_gemm", "w8_a16_gemm_", "preprocess_weights", "quant_weights"]

build/torch26-cxx11-cu126-aarch64-linux/quantization_eetq/_ops.py

@@ -0,0 +1,9 @@
+import torch
+from . import _quantization_eetq_f87291e
+ops = torch.ops._quantization_eetq_f87291e
+
+def add_op_namespace_prefix(op_name: str):
+    """
+    Prefix op by namespace.
+    """
+    return f"_quantization_eetq_f87291e::{op_name}"

build/torch26-cxx11-cu126-aarch64-linux/quantization_eetq/_quantization_eetq_f87291e.abi3.so

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2b8b5343121a6ce83d7730e31913ed787ce1ebacd7bfb562ac9bab5fcbd4f7e
+size 26495968

build/torch26-cxx11-cu126-aarch64-linux/quantization_eetq/custom_ops.py

@@ -0,0 +1,36 @@
+from typing import List
+import torch
+
+from ._ops import ops
+
+
+def w8_a16_gemm(
+    input: torch.Tensor, weight: torch.Tensor, scale: torch.Tensor
+) -> torch.Tensor:
+    return ops.w8_a16_gemm(input, weight, scale)
+
+
+def w8_a16_gemm_(
+    input: torch.Tensor,
+    weight: torch.Tensor,
+    scale: torch.Tensor,
+    output: torch.Tensor,
+    m: int,
+    n: int,
+    k: int,
+) -> torch.Tensor:
+    return ops.w8_a16_gemm_(input, weight, scale, output, m, n, k)
+
+
+def preprocess_weights(origin_weight: torch.Tensor, is_int4: bool) -> torch.Tensor:
+    return ops.preprocess_weights(origin_weight, is_int4)
+
+
+def quant_weights(
+    origin_weight: torch.Tensor,
+    quant_type: torch.dtype,
+    return_unprocessed_quantized_tensor: bool,
+) -> List[torch.Tensor]:
+    return ops.quant_weights(
+        origin_weight, quant_type, return_unprocessed_quantized_tensor
+    )

build/torch26-cxx98-cu126-aarch64-linux/quantization_eetq/__init__.py

@@ -0,0 +1,3 @@
+from .custom_ops import w8_a16_gemm, w8_a16_gemm_, preprocess_weights, quant_weights
+
+__all__ = ["w8_a16_gemm", "w8_a16_gemm_", "preprocess_weights", "quant_weights"]

build/torch26-cxx98-cu126-aarch64-linux/quantization_eetq/_ops.py

@@ -0,0 +1,9 @@
+import torch
+from . import _quantization_eetq_f87291e
+ops = torch.ops._quantization_eetq_f87291e
+
+def add_op_namespace_prefix(op_name: str):
+    """
+    Prefix op by namespace.
+    """
+    return f"_quantization_eetq_f87291e::{op_name}"

build/torch26-cxx98-cu126-aarch64-linux/quantization_eetq/_quantization_eetq_f87291e.abi3.so

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c0c6116e6e6fb38a5925a055966845b52920ca58e77f8a2bc097d22d82a45723
+size 26757632

build/torch26-cxx98-cu126-aarch64-linux/quantization_eetq/custom_ops.py

@@ -0,0 +1,36 @@
+from typing import List
+import torch
+
+from ._ops import ops
+
+
+def w8_a16_gemm(
+    input: torch.Tensor, weight: torch.Tensor, scale: torch.Tensor
+) -> torch.Tensor:
+    return ops.w8_a16_gemm(input, weight, scale)
+
+
+def w8_a16_gemm_(
+    input: torch.Tensor,
+    weight: torch.Tensor,
+    scale: torch.Tensor,
+    output: torch.Tensor,
+    m: int,
+    n: int,
+    k: int,
+) -> torch.Tensor:
+    return ops.w8_a16_gemm_(input, weight, scale, output, m, n, k)
+
+
+def preprocess_weights(origin_weight: torch.Tensor, is_int4: bool) -> torch.Tensor:
+    return ops.preprocess_weights(origin_weight, is_int4)
+
+
+def quant_weights(
+    origin_weight: torch.Tensor,
+    quant_type: torch.dtype,
+    return_unprocessed_quantized_tensor: bool,
+) -> List[torch.Tensor]:
+    return ops.quant_weights(
+        origin_weight, quant_type, return_unprocessed_quantized_tensor
+    )

build/torch27-cxx11-cu126-aarch64-linux/quantization_eetq/__init__.py

@@ -0,0 +1,3 @@
+from .custom_ops import w8_a16_gemm, w8_a16_gemm_, preprocess_weights, quant_weights
+
+__all__ = ["w8_a16_gemm", "w8_a16_gemm_", "preprocess_weights", "quant_weights"]

build/torch27-cxx11-cu126-aarch64-linux/quantization_eetq/_ops.py

@@ -0,0 +1,9 @@
+import torch
+from . import _quantization_eetq_f87291e
+ops = torch.ops._quantization_eetq_f87291e
+
+def add_op_namespace_prefix(op_name: str):
+    """
+    Prefix op by namespace.
+    """
+    return f"_quantization_eetq_f87291e::{op_name}"

build/torch27-cxx11-cu126-aarch64-linux/quantization_eetq/_quantization_eetq_f87291e.abi3.so

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a73dd746900ddb1570192c2cb50736aa56f8854a1dcd0d84c746138217cd82b
+size 26496176

build/torch27-cxx11-cu126-aarch64-linux/quantization_eetq/custom_ops.py

@@ -0,0 +1,36 @@
+from typing import List
+import torch
+
+from ._ops import ops
+
+
+def w8_a16_gemm(
+    input: torch.Tensor, weight: torch.Tensor, scale: torch.Tensor
+) -> torch.Tensor:
+    return ops.w8_a16_gemm(input, weight, scale)
+
+
+def w8_a16_gemm_(
+    input: torch.Tensor,
+    weight: torch.Tensor,
+    scale: torch.Tensor,
+    output: torch.Tensor,
+    m: int,
+    n: int,
+    k: int,
+) -> torch.Tensor:
+    return ops.w8_a16_gemm_(input, weight, scale, output, m, n, k)
+
+
+def preprocess_weights(origin_weight: torch.Tensor, is_int4: bool) -> torch.Tensor:
+    return ops.preprocess_weights(origin_weight, is_int4)
+
+
+def quant_weights(
+    origin_weight: torch.Tensor,
+    quant_type: torch.dtype,
+    return_unprocessed_quantized_tensor: bool,
+) -> List[torch.Tensor]:
+    return ops.quant_weights(
+        origin_weight, quant_type, return_unprocessed_quantized_tensor
+    )

build/torch27-cxx11-cu128-aarch64-linux/quantization_eetq/__init__.py

@@ -0,0 +1,3 @@
+from .custom_ops import w8_a16_gemm, w8_a16_gemm_, preprocess_weights, quant_weights
+
+__all__ = ["w8_a16_gemm", "w8_a16_gemm_", "preprocess_weights", "quant_weights"]

build/torch27-cxx11-cu128-aarch64-linux/quantization_eetq/_ops.py

@@ -0,0 +1,9 @@
+import torch
+from . import _quantization_eetq_f87291e
+ops = torch.ops._quantization_eetq_f87291e
+
+def add_op_namespace_prefix(op_name: str):
+    """
+    Prefix op by namespace.
+    """
+    return f"_quantization_eetq_f87291e::{op_name}"

build/torch27-cxx11-cu128-aarch64-linux/quantization_eetq/_quantization_eetq_f87291e.abi3.so

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4d40830eefb300f3c6e042e68baccf936811f3ea6de26f54d26b602b96e6e29
+size 27936256

build/torch27-cxx11-cu128-aarch64-linux/quantization_eetq/custom_ops.py

@@ -0,0 +1,36 @@
+from typing import List
+import torch
+
+from ._ops import ops
+
+
+def w8_a16_gemm(
+    input: torch.Tensor, weight: torch.Tensor, scale: torch.Tensor
+) -> torch.Tensor:
+    return ops.w8_a16_gemm(input, weight, scale)
+
+
+def w8_a16_gemm_(
+    input: torch.Tensor,
+    weight: torch.Tensor,
+    scale: torch.Tensor,
+    output: torch.Tensor,
+    m: int,
+    n: int,
+    k: int,
+) -> torch.Tensor:
+    return ops.w8_a16_gemm_(input, weight, scale, output, m, n, k)
+
+
+def preprocess_weights(origin_weight: torch.Tensor, is_int4: bool) -> torch.Tensor:
+    return ops.preprocess_weights(origin_weight, is_int4)
+
+
+def quant_weights(
+    origin_weight: torch.Tensor,
+    quant_type: torch.dtype,
+    return_unprocessed_quantized_tensor: bool,
+) -> List[torch.Tensor]:
+    return ops.quant_weights(
+        origin_weight, quant_type, return_unprocessed_quantized_tensor
+    )