HK
Heykuki News
Top
New
Best
Ask
Show
Jobs
Toggle theme
Top
New
Best
Ask
Show
Jobs
Request
1.
▲
Fp8 runs ~100 tflops faster when the kernel name has "cutlass" in it
github.com/triton-lang
166 comments
9 months ago
mmastrac
338 points
2.
▲
Gluon: a GPU programming language based on the same compiler stack as Triton
github.com/triton-lang
24 comments
9 months ago
matt_d
83 points
3.
▲
Fp8 runs ~100 tflops faster when the kernel name has "cutlass" in it
github.com/triton-lang
discuss
a year ago
mmastrac
4 points
4.
▲
Triton Extensions: a framework for developing and building compiler extensions
github.com/triton-lang
discuss
5 months ago
matt_d
2 points
5.
▲
Triton Plugins
github.com/triton-lang
discuss
7 months ago
zer0zzz
2 points
6.
▲
Triton Support for Blackwell
github.com/triton-lang
discuss
a year ago
elgatolopez
2 points
7.
▲
I fixed a segfault in Triton that broke every RTX 5070/5080/5090
github.com/triton-lang
discuss
3 months ago
pat90000
1 points
8.
▲
Triton CUDA Tile IR Back End
github.com/triton-lang
discuss
5 months ago
my123
1 points
9.
▲
Automatic Warp Specialization in Triton
github.com/triton-lang
discuss
a year ago
subharmonicon
1 points
10.
▲
Triton Language and Compiler
github.com/openai
discuss
3 years ago
tosh
3 points
11.
▲
OpenAI Triton: language and compiler for highly efficient Deep-Learning
github.com/openai
discuss
2 years ago
tosh
1 points
12.
▲
Show HN: 80% faster, 50% less memory, 0% loss of accuracy Llama finetuning
github.com/unslothai
119 comments
3 years ago
danielhanchen
385 points
13.
▲
Show HN: Finetune Llama-3.1 2x faster in a Colab
colab.research.google.com
2 comments
2 years ago
danielhanchen
16 points
14.
▲
Finetune language models 30x faster
unsloth.ai
discuss
3 years ago
danielhanchen
2 points
15.
▲
Show HN: Efficient `Torch.cdist` Using Triton
github.com/jinensetpal
discuss
a year ago
codeinassembly
1 points