Promote latest kernel artifacts to main

README.md

@@ -1,9 +1,21 @@
-# flashrt/grouped-moe-gemv
+---
+library_name: kernels
+license: apache-2.0
+tags:
+  - cuda
+  - native-cuda
+  - flashrt
+  - moe
+  - nvfp4
+  - blackwell
+---
 
-This repository is a compatibility mirror for older `kernels` clients
-that resolve repositories through the default Hugging Face model repo API.
+# grouped-moe-gemv
 
-Canonical Kernel Hub repo: https://huggingface.co/kernels/flashrt/grouped-moe-gemv
+Native CUDA FlashRT grouped MoE GEMV kernels for BF16 activations and NVFP4
+weights.
 
-Do not edit this mirror by hand. It is generated from the Kernel Hub
-`vN` branches and contains the same `build/**` artifacts.
+Available functions:
+
+- `w4a16_decode_gemv_bf16`
+- `grouped_w4a16_gemv_bf16`

build/torch211-cxx11-cu128-x86_64-linux/{_grouped_moe_gemv_cuda_4baeadf.abi3.so → _grouped_moe_gemv_cuda_ca580d0.abi3.so}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0f6ea7ddd3e3857122d22639a2dcb781810da11812dbfa1d28e4cd7a4aba8edf
+oid sha256:c354d06796eee85f3f5881ffd80da2c25e5cb19ca79e4022cedc32365db929db
 size 309656

build/torch211-cxx11-cu128-x86_64-linux/_ops.py

@@ -1,9 +1,9 @@
 import torch
-from . import _grouped_moe_gemv_cuda_4baeadf
-ops = torch.ops._grouped_moe_gemv_cuda_4baeadf
+from . import _grouped_moe_gemv_cuda_ca580d0
+ops = torch.ops._grouped_moe_gemv_cuda_ca580d0
 
 def add_op_namespace_prefix(op_name: str):
     """
     Prefix op by namespace.
     """
-    return f"_grouped_moe_gemv_cuda_4baeadf::{op_name}"
+    return f"_grouped_moe_gemv_cuda_ca580d0::{op_name}"

build/torch211-cxx11-cu128-x86_64-linux/metadata.json

@@ -1,6 +1,6 @@
 {
   "name": "grouped-moe-gemv",
-  "id": "_grouped_moe_gemv_cuda_4baeadf",
+  "id": "_grouped_moe_gemv_cuda_ca580d0",
   "version": 2,
   "license": "Apache-2.0",
   "python-depends": [],

build/torch211-cxx11-cu130-x86_64-linux/{_grouped_moe_gemv_cuda_4baeadf.abi3.so → _grouped_moe_gemv_cuda_ca580d0.abi3.so}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9229ff3ee2c5cd0ff821f9749690aeefe19d6e56205a895192a6f7e39eb944e3
+oid sha256:68d6125ca405f8cb8a69a1b3bdc49dfaaa40470137dc9676a74891d084e71e7f
 size 303728

build/torch211-cxx11-cu130-x86_64-linux/_ops.py

@@ -1,9 +1,9 @@
 import torch
-from . import _grouped_moe_gemv_cuda_4baeadf
-ops = torch.ops._grouped_moe_gemv_cuda_4baeadf
+from . import _grouped_moe_gemv_cuda_ca580d0
+ops = torch.ops._grouped_moe_gemv_cuda_ca580d0
 
 def add_op_namespace_prefix(op_name: str):
     """
     Prefix op by namespace.
     """
-    return f"_grouped_moe_gemv_cuda_4baeadf::{op_name}"
+    return f"_grouped_moe_gemv_cuda_ca580d0::{op_name}"

build/torch211-cxx11-cu130-x86_64-linux/metadata.json

@@ -1,6 +1,6 @@
 {
   "name": "grouped-moe-gemv",
-  "id": "_grouped_moe_gemv_cuda_4baeadf",
+  "id": "_grouped_moe_gemv_cuda_ca580d0",
   "version": 2,
   "license": "Apache-2.0",
   "python-depends": [],

build/torch212-cxx11-cu130-x86_64-linux/{_grouped_moe_gemv_cuda_4baeadf.abi3.so → _grouped_moe_gemv_cuda_ca580d0.abi3.so}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9861c6db3fbca4988bb4b5e7c48a6850c2d558a71a50c4fadaadb615f9b1fe5b
+oid sha256:b5d24206f7787e86cf5476f41d9f28ebd060345c8682c56341545aa6ff31bea3
 size 314400

build/torch212-cxx11-cu130-x86_64-linux/_ops.py

@@ -1,9 +1,9 @@
 import torch
-from . import _grouped_moe_gemv_cuda_4baeadf
-ops = torch.ops._grouped_moe_gemv_cuda_4baeadf
+from . import _grouped_moe_gemv_cuda_ca580d0
+ops = torch.ops._grouped_moe_gemv_cuda_ca580d0
 
 def add_op_namespace_prefix(op_name: str):
     """
     Prefix op by namespace.
     """
-    return f"_grouped_moe_gemv_cuda_4baeadf::{op_name}"
+    return f"_grouped_moe_gemv_cuda_ca580d0::{op_name}"

build/torch212-cxx11-cu130-x86_64-linux/metadata.json

@@ -1,6 +1,6 @@
 {
   "name": "grouped-moe-gemv",
-  "id": "_grouped_moe_gemv_cuda_4baeadf",
+  "id": "_grouped_moe_gemv_cuda_ca580d0",
   "version": 2,
   "license": "Apache-2.0",
   "python-depends": [],

build/torch212-cxx11-cu132-x86_64-linux/{_grouped_moe_gemv_cuda_4baeadf.abi3.so → _grouped_moe_gemv_cuda_ca580d0.abi3.so}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:547f5e5a75fdc8b297c23baecdfaad9391bd82361046435cb13b38028a36f45b
+oid sha256:1d558eb2222a447fe44cbc5e0f95d62663264f7ec2115804a97e05a96ddc3f23
 size 318552

build/torch212-cxx11-cu132-x86_64-linux/_ops.py

@@ -1,9 +1,9 @@
 import torch
-from . import _grouped_moe_gemv_cuda_4baeadf
-ops = torch.ops._grouped_moe_gemv_cuda_4baeadf
+from . import _grouped_moe_gemv_cuda_ca580d0
+ops = torch.ops._grouped_moe_gemv_cuda_ca580d0
 
 def add_op_namespace_prefix(op_name: str):
     """
     Prefix op by namespace.
     """
-    return f"_grouped_moe_gemv_cuda_4baeadf::{op_name}"
+    return f"_grouped_moe_gemv_cuda_ca580d0::{op_name}"

build/torch212-cxx11-cu132-x86_64-linux/metadata.json

@@ -1,6 +1,6 @@
 {
   "name": "grouped-moe-gemv",
-  "id": "_grouped_moe_gemv_cuda_4baeadf",
+  "id": "_grouped_moe_gemv_cuda_ca580d0",
   "version": 2,
   "license": "Apache-2.0",
   "python-depends": [],