English | 简体中文

Flash-DMA is a high-performance attention implementation that integrates Flash Attention's memory efficiency with Dynamic Mask Attention's sparse computation capabilities for processing extremely long sequences in transformer models.

Key Features

• Dynamic Sparse Attention: Dynamically selects the most relevant keys for each query, reducing computational complexity from $O(N^2)$ to $O(N \cdot w)$ where $w \ll N$, supporting trainable sparse patterns.
• Memory Efficiency: Maintains Flash Attention's $O(N)$ memory complexity without instantiating the full attention matrix.
• CUDA Deep Optimization: Utilizes custom CUDA kernels with shared memory aliasing, pipelined prefetching, and block skipping for high throughput and low memory access overhead.
• Extremely Long Context Support: Handles 128K+ token sequences efficiently through dynamic mask windowing while preserving accuracy.
• Learnable Bias: Built-in learnable attention bias and its gradient path dbias, eliminating the need for additional external operators.
• Fusion-Friendly Training: Both forward and backward passes support block-level zero-mask skipping, further reducing computation in sparse scenarios.

Performance

We present expected speedup of Flash-DMA over standard PyTorch SDPA.

Installation

Prerequisites

• Python: 3.8 or later
• PyTorch: 2.0.0 or later
• CUDA: 11.8 or later
• NVIDIA GPU: Compute Capability 8.0 or higher
• C++ Compiler: GCC 7+

CUDA Environment Setup

Ensure your CUDA environment is properly configured:

# Check CUDA installation
nvcc --version

# Set CUDA_HOME if needed
export CUDA_HOME=/usr/local/cuda

Install from Source

git clone https://github.com/SmallDoges/flash-dmattn.git
cd flash-dmattn
git submodule update --init --recursive
pip install .

Quick Start

import torch
from flash_dmattn import flash_dmattn_func_auto
import math

# Setup
batch_size, seq_len, num_heads, head_dim = 2, 4096, 16, 128
device = torch.device('cuda')
dtype = torch.bfloat16

# Input tensors
query = torch.randn(batch_size, seq_len, num_heads, head_dim, 
                   device=device, dtype=dtype)
key = torch.randn(batch_size, seq_len, num_heads, head_dim,
                 device=device, dtype=dtype)
value = torch.randn(batch_size, seq_len, num_heads, head_dim,
                   device=device, dtype=dtype)

# Create mask and bias for sparse attention
attention_bias = torch.randn(batch_size, num_heads, seq_len, seq_len,
                           device=device, dtype=dtype)
attention_mask = torch.ones(batch_size, num_heads, seq_len, seq_len,
                          device=device, dtype=dtype)

# Apply dynamic masking (keep top-k for long sequences)
keep_window_size = 2048
if seq_len > keep_window_size:
    # Select top-k most important keys for each query
    topk_indices = torch.topk(attention_bias, keep_window_size, dim=-1, 
                             largest=True, sorted=False).indices
    attention_mask.zero_()
    attention_mask.scatter(-1, topk_indices, 1.0)

# Select backend
flash_dmattn_func = flash_dmattn_func_auto(backend="cuda")

# Run Flash Dynamic Mask Attention
output = flash_dmattn_func(
    q=query,
    k=key, 
    v=value,
    attn_mask=attention_mask,
    attn_bias=attention_bias,
    is_causal=True,
    scale=1.0/math.sqrt(head_dim),
)

print(f"Output shape: {output.shape}")  # [2, 4096, 16, 128]

How It Works

Flash-DMA combines two complementary techniques:

• Dynamic Mask Attention: Computes relevance scores for keys and selects only the most important ones for attention computation
• Flash Attention: Processes attention in blocks to reduce memory usage and HBM access

The Integration Approach

The integration happens at the CUDA kernel level with several key components:

• ZOH States: Pre-computed importance scores for key selection
• Active Masks: Binary masks indicating which keys should be considered for each query
• Sparse Skipping: Custom CUDA kernels for efficient sparse attention computation
• Block-Based Processing: Maintains Flash Attention's block-based approach for memory efficiency

This creates a hybrid attention mechanism that achieves both memory and computational efficiency for long sequences.

Documentation

📚 Complete documentation is available in the docs directory:

• API Reference - Complete function documentation and usage examples
• Integration Guide - Detailed technical documentation of the Flash Attention integration

Building from Source

Development Setup

# Clone with submodules
git clone --recursive https://github.com/SmallDoges/flash-dmattn.git
cd flash-dmattn

# Build in development mode
pip install -e .

# Run tests to verify installation
python -c "import flash_dma_cuda; print('✅ Flash DMA CUDA extension imported successfully')"

Build Requirements

• CUDA Toolkit 11.8+
• CUTLASS library
• PyTorch with CUDA support

Supported Architectures

• SM 8.0
• SM 9.0
• SM 10.0
• SM 12.0

Note: Flash Dynamic Mask Attention requires CUDA compute capability 8.0+ for optimal performance. Earlier architectures are not supported.

Benchmarking

Flash-DMA provides comprehensive benchmarking tools to evaluate performance across different configurations:

Forward Pass Equivalence

python benchmarks/forward_equivalence.py

Validates numerical consistency between Python reference and CUDA implementation.

Forward Pass Performance Benchmarking

python benchmarks/forward_performance.py

Compares Flash-DMA against standard SDPA across various sequence lengths and batch sizes.

Backward Pass Equivalence

python benchmarks/backward_equivalence.py

Validates numerical consistency between Python reference and CUDA implementation.

Backward Pass Performance Benchmarking

python benchmarks/backward_performance.py

Compares Flash-DMA against standard SDPA across various sequence lengths and batch sizes.

Gradient Computation

python benchmarks/grad_equivalence.py

Tests backward pass implementation and gradient equivalence.

Troubleshooting

Common Issues

Compilation Errors

# Ensure CUDA_HOME is set correctly
echo $CUDA_HOME  # Linux/Mac
echo $env:CUDA_HOME  # Windows PowerShell

# Check CUDA toolkit version
nvcc --version

# Verify PyTorch CUDA support
python -c "import torch; print(f'CUDA available: {torch.cuda.is_available()}')"

Import Errors

# Test basic import
try:
    from flash_dmattn import flash_dmattn_func, get_available_backends
    print("✅ Flash Dynamic Mask Attention imported successfully")
    print(f"Available backends: {get_available_backends()}")
except ImportError as e:
    print(f"❌ Import failed: {e}")
    print("Please ensure the package is properly installed with: pip install -e .")

Performance Issues

# Monitor GPU memory usage
from flash_dmattn import flash_dmattn_func

def print_memory_stats():
    if torch.cuda.is_available():
        print(f"GPU Memory: {torch.cuda.memory_allocated() / 1e9:.2f} GB")

print_memory_stats()
output = flash_dmattn_func(q=query, k=key, v=value, is_causal=True)
print_memory_stats()

# Clear cache if needed
torch.cuda.empty_cache()

Contributing

We welcome contributions from the community! Flash-DMA is an open-source project and we value all types of contributions.

How to Contribute

• Report bugs: Found a bug? Please open an issue
• Request features: Have an idea for improvement? Let us know
• Submit code: Ready to contribute code? Check our Contributing Guide
• Improve docs: Help us make the documentation better

Quick Start for Contributors

1. Fork the repository
2. Create a feature branch: git checkout -b feature-name
3. Make your changes and test them
4. Submit a pull request

For detailed instructions, see our Contributing Guide.

Code of Conduct

This project follows the Contributor Covenant Code of Conduct. By participating, you are expected to uphold this code.

License

This project is licensed under the BSD 3-Clause License. See LICENSE for details.

Citation

If you use Flash-DMA in your research, please cite:

@misc{shi2025trainabledynamicmasksparse,
      title={Trainable Dynamic Mask Sparse Attention}, 
      author={Jingze Shi and Yifan Wu and Bingheng Wu and Yiran Peng and Liangdong Wang and Guang Liu and Yuyu Luo},
      year={2025},
      eprint={2508.02124},
      archivePrefix={arXiv},
      primaryClass={cs.AI},
      url={https://arxiv.org/abs/2508.02124}, 
}

Acknowledgments

This project builds upon and integrates several excellent works:

• OpenSeek - Kernel development support
• Flash-Attention - Memory-efficient attention computation
• NVIDIA CUTLASS - High-performance matrix operations library

We thank the open-source community for their contributions to efficient transformer implementations. 🤗