Hugging Face's logo Hugging Face
  • Models
  • Datasets
  • Spaces
  • Buckets new
  • Docs
  • Enterprise
  • Pricing
    • Website
      • Tasks
      • HuggingChat
      • Collections
      • Languages
      • Organizations
    • Community
      • Blog
      • Posts
      • Daily Papers
      • Hardware
      • Learn
      • Discord
      • Forum
      • GitHub
    • Solutions
      • Team & Enterprise
      • Hugging Face PRO
      • Enterprise Support
      • Inference Providers
      • Inference Endpoints
      • Storage Buckets

  • Log In
  • Sign Up

jburtoft
/
kda-neuron-kernels

Kernels
kernel
neuron
trainium
kda
linear-attention
fla-core
training
backward
Model card Files Files and versions
xet
Community

Instructions to use jburtoft/kda-neuron-kernels with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

  • Libraries
  • Kernels

    How to use jburtoft/kda-neuron-kernels with Kernels:

    # !pip install kernels
    
    from kernels import get_kernel
    
    kernel = get_kernel("jburtoft/kda-neuron-kernels")
  • Notebooks
  • Google Colab
  • Kaggle
kda-neuron-kernels / build /torch-neuron
185 kB
Ctrl+K
Ctrl+K
  • 1 contributor
History: 4 commits
jburtoft's picture
jburtoft
v1.3: fused multi-chunk backward (kda_chunked_fused, 2-5x faster training)
6a4016a verified about 10 hours ago
  • __init__.py
    8.26 kB
    v1.3: fused multi-chunk backward (kda_chunked_fused, 2-5x faster training) about 10 hours ago
  • kda_autograd.py
    12.4 kB
    v1.3: fused multi-chunk backward (kda_chunked_fused, 2-5x faster training) about 10 hours ago
  • kda_chunk_bwd.py
    30 kB
    v1.2: training / backward support (kda_recurrent, kda_chunked + backward kernels) about 11 hours ago
  • kda_chunk_step_v2.py
    20.2 kB
    v1.2: training / backward support (kda_recurrent, kda_chunked + backward kernels) about 11 hours ago
  • kda_fused_chunked_bwd.py
    31.7 kB
    v1.3: fused multi-chunk backward (kda_chunked_fused, 2-5x faster training) about 10 hours ago
  • kda_fused_chunked_fwd.py
    21 kB
    v1.3: fused multi-chunk backward (kda_chunked_fused, 2-5x faster training) about 10 hours ago
  • kda_recurrent_bwd.py
    16.5 kB
    v1.2: training / backward support (kda_recurrent, kda_chunked + backward kernels) about 11 hours ago
  • kda_recurrent_fwd_v2.py
    6.56 kB
    v1.2: training / backward support (kda_recurrent, kda_chunked + backward kernels) about 11 hours ago
  • nki_kda.py
    14.8 kB
    v1.1: T-KDA-02 fix + fully model-agnostic (remove all consumer-model references) about 18 hours ago
  • nki_kda_chunked.py
    23.4 kB
    v1.1: T-KDA-02 fix + fully model-agnostic (remove all consumer-model references) about 18 hours ago