Downloads · 30 days
0
triton-gdn-decode
baladhurgesh97/triton-gdn-decode
triton-gdn-decode is a machine learning model from baladhurgesh97. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for kernels. The card lists the license as apache-2.0.
Triton packed Gated DeltaNet decode kernel. One CTA per head loops over V-tiles (BV=32, numwarps=4) instead of launching NV=4 CTAs per head. Tuned on NVIDIA GB10 for AxionML/Qwen3.5-0.8B-NVFP4.
Access
Public
Updated Aug 22, 2026
Repo size
—
Likes
1
Public
Repo makeup
Click a slice to open those files.
.py14.9 KB · 79%
At a glance
- Library
- kernels
- License
- apache-2.0
- Access
- Public
- Created
- Aug 22, 2026
- Updated
- Aug 22, 2026
- SHA
- eddc5f09
Model card
From the Hugging Face model README
gdn-packed-decode
Triton packed Gated DeltaNet decode kernel. One CTA per head loops
over V-tiles (BV=32, num_warps=4) instead of launching NV=4 CTAs
per head. Tuned on NVIDIA GB10 for
AxionML/Qwen3.5-0.8B-NVFP4.
Weights are unchanged. This is a kernel repo, not a model checkpoint.
Load from the Hub
from kernels import get_kernel
k = get_kernel("<this-repo>", trust_remote_code=True)
# k.fused_recurrent_gated_delta_rule_packed_decode(...)
trust_remote_code=True is required unless the publisher is on the
kernels trusted list. Replace <this-repo> with this repository id.
Patch into vLLM
hf download <this-repo> --local-dir ./gdn-packed-decode
python3 ./gdn-packed-decode/apply_kernel.py apply
# restart vLLM so EngineCore reimports, then:
# vllm serve AxionML/Qwen3.5-0.8B-NVFP4
GSM8K-10 (GB10, vLLM 0.27.1, frozen indices, thinking off)
| Kernel | Acc | Median decode tok/s |
|---|---|---|
| Upstream vLLM | 5/10 | 205.45 |
| This kernel (run 1) | 5/10 | 219.90 |
| This kernel (run 2) | 5/10 | 219.89 |
+7.0% e2e. Correctness vs upstream FLA: 6/6 (atol=rtol=2e-2).
Functions
fused_recurrent_gated_delta_rule_packed_decode
Same signature as vLLM FLA packed decode. Returns (out, initial_state).
- Library
- kernels
- License
- apache-2.0
- Created
- Aug 22, 2026
- Updated
- Aug 22, 2026