Uploaded using `kernel-builder`.

build/torch210-cxx11-cu126-aarch64-linux/_ops.py

@@ -1,9 +1,9 @@
 import torch
-from . import _quantization_eetq_cuda_86f75d9
-ops = torch.ops._quantization_eetq_cuda_86f75d9
+from . import _quantization_eetq_cuda_2d6119c
+ops = torch.ops._quantization_eetq_cuda_2d6119c
 
 def add_op_namespace_prefix(op_name: str):
     """
     Prefix op by namespace.
     """
-    return f"_quantization_eetq_cuda_86f75d9::{op_name}"
+    return f"_quantization_eetq_cuda_2d6119c::{op_name}"

build/torch210-cxx11-cu126-aarch64-linux/{_quantization_eetq_cuda_86f75d9.abi3.so → _quantization_eetq_cuda_2d6119c.abi3.so}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3c76662fb7202c7f7410dbfcaf69e8b92c77545859171d26481d2d37b27201a4
+oid sha256:40d168ca6634a23dded922ae1bd52d8e1a2ad30701de31e7e54c826979cf5512
 size 39010048

build/torch210-cxx11-cu126-aarch64-linux/metadata.json

@@ -1,6 +1,6 @@
 {
   "name": "quantization-eetq",
-  "id": "_quantization_eetq_cuda_86f75d9",
+  "id": "_quantization_eetq_cuda_2d6119c",
   "version": 1,
   "license": "Apache-2.0",
   "python-depends": [],

build/torch210-cxx11-cu128-aarch64-linux/_ops.py

@@ -1,9 +1,9 @@
 import torch
-from . import _quantization_eetq_cuda_86f75d9
-ops = torch.ops._quantization_eetq_cuda_86f75d9
+from . import _quantization_eetq_cuda_2d6119c
+ops = torch.ops._quantization_eetq_cuda_2d6119c
 
 def add_op_namespace_prefix(op_name: str):
     """
     Prefix op by namespace.
     """
-    return f"_quantization_eetq_cuda_86f75d9::{op_name}"
+    return f"_quantization_eetq_cuda_2d6119c::{op_name}"

build/torch210-cxx11-cu128-aarch64-linux/{_quantization_eetq_cuda_86f75d9.abi3.so → _quantization_eetq_cuda_2d6119c.abi3.so}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c88c95f4b6331e387278fbbf05c0dd7f39efd6bc262ef34c9bdc63b40f53489f
+oid sha256:4f185271e38f588a714e8b360fc7491df3a9592c330c58e805879767f05f92d0
 size 45366048

build/torch210-cxx11-cu128-aarch64-linux/metadata.json

@@ -1,6 +1,6 @@
 {
   "name": "quantization-eetq",
-  "id": "_quantization_eetq_cuda_86f75d9",
+  "id": "_quantization_eetq_cuda_2d6119c",
   "version": 1,
   "license": "Apache-2.0",
   "python-depends": [],

build/torch211-cxx11-cu126-aarch64-linux/_ops.py

@@ -1,9 +1,9 @@
 import torch
-from . import _quantization_eetq_cuda_86f75d9
-ops = torch.ops._quantization_eetq_cuda_86f75d9
+from . import _quantization_eetq_cuda_2d6119c
+ops = torch.ops._quantization_eetq_cuda_2d6119c
 
 def add_op_namespace_prefix(op_name: str):
     """
     Prefix op by namespace.
     """
-    return f"_quantization_eetq_cuda_86f75d9::{op_name}"
+    return f"_quantization_eetq_cuda_2d6119c::{op_name}"

build/torch211-cxx11-cu126-aarch64-linux/{_quantization_eetq_cuda_86f75d9.abi3.so → _quantization_eetq_cuda_2d6119c.abi3.so}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e09ba83a7a2cf4bad5120571b73251e2fe54097c03b13a356d674fcf25108c1c
+oid sha256:9166428182ff8357a8124b5fa9b45557867419cd7095c16264c0d7b4a8b1985c
 size 39006256

build/torch211-cxx11-cu126-aarch64-linux/metadata.json

@@ -1,6 +1,6 @@
 {
   "name": "quantization-eetq",
-  "id": "_quantization_eetq_cuda_86f75d9",
+  "id": "_quantization_eetq_cuda_2d6119c",
   "version": 1,
   "license": "Apache-2.0",
   "python-depends": [],

build/torch211-cxx11-cu128-aarch64-linux/_ops.py

@@ -1,9 +1,9 @@
 import torch
-from . import _quantization_eetq_cuda_86f75d9
-ops = torch.ops._quantization_eetq_cuda_86f75d9
+from . import _quantization_eetq_cuda_2d6119c
+ops = torch.ops._quantization_eetq_cuda_2d6119c
 
 def add_op_namespace_prefix(op_name: str):
     """
     Prefix op by namespace.
     """
-    return f"_quantization_eetq_cuda_86f75d9::{op_name}"
+    return f"_quantization_eetq_cuda_2d6119c::{op_name}"

build/torch211-cxx11-cu128-aarch64-linux/{_quantization_eetq_cuda_86f75d9.abi3.so → _quantization_eetq_cuda_2d6119c.abi3.so}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:74fac81d793d214112fc6f20f13cd69dd64e745e20d88a6807c6da033646d6da
+oid sha256:e1eb5bfd99bc0c2ad9f430b85f50d5e5075103bd63f0561bf2678472d7101e97
 size 45296712

build/torch211-cxx11-cu128-aarch64-linux/metadata.json

@@ -1,6 +1,6 @@
 {
   "name": "quantization-eetq",
-  "id": "_quantization_eetq_cuda_86f75d9",
+  "id": "_quantization_eetq_cuda_2d6119c",
   "version": 1,
   "license": "Apache-2.0",
   "python-depends": [],

build/torch212-cxx11-cu126-aarch64-linux/__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/torch212-cxx11-cu126-aarch64-linux/_ops.py

@@ -0,0 +1,9 @@
+import torch
+from . import _quantization_eetq_cuda_2d6119c
+ops = torch.ops._quantization_eetq_cuda_2d6119c
+
+def add_op_namespace_prefix(op_name: str):
+    """
+    Prefix op by namespace.
+    """
+    return f"_quantization_eetq_cuda_2d6119c::{op_name}"

build/torch212-cxx11-cu126-aarch64-linux/_quantization_eetq_cuda_2d6119c.abi3.so

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9fb6c323866edae12a279ad9d01ebe1565424693e9cb5b8e4713bd8e72d09818
+size 39009824

build/torch212-cxx11-cu126-aarch64-linux/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/torch212-cxx11-cu126-aarch64-linux/metadata.json

@@ -0,0 +1,20 @@
+{
+  "name": "quantization-eetq",
+  "id": "_quantization_eetq_cuda_2d6119c",
+  "version": 1,
+  "license": "Apache-2.0",
+  "python-depends": [],
+  "backend": {
+    "type": "cuda",
+    "archs": [
+      "7.0",
+      "7.2",
+      "7.5",
+      "8.0",
+      "8.6",
+      "8.7",
+      "8.9",
+      "9.0+PTX"
+    ]
+  }
+}

build/torch212-cxx11-cu126-aarch64-linux/quantization_eetq/__init__.py

@@ -0,0 +1,26 @@
+import ctypes
+import importlib.util
+import sys
+from pathlib import Path
+from types import ModuleType
+
+
+def _import_from_path(file_path: Path) -> ModuleType:
+    # We cannot use the module name as-is, after adding it to `sys.modules`,
+    # it would also be used for other imports. So, we make a module name that
+    # depends on the path for it to be unique using the hex-encoded hash of
+    # the path.
+    path_hash = "{:x}".format(ctypes.c_size_t(hash(file_path.absolute())).value)
+    module_name = path_hash
+    spec = importlib.util.spec_from_file_location(module_name, file_path)
+    if spec is None:
+        raise ImportError(f"Cannot load spec for {module_name} from {file_path}")
+    module = importlib.util.module_from_spec(spec)
+    if module is None:
+        raise ImportError(f"Cannot load module {module_name} from spec")
+    sys.modules[module_name] = module
+    spec.loader.exec_module(module)  # type: ignore
+    return module
+
+
+globals().update(vars(_import_from_path(Path(__file__).parent.parent / "__init__.py")))