This is stars and forks stats for /philipturner/metal-flash-attention repository. As of 29 Apr, 2024 this repository has 169 stars and 4 forks.
Metal FlashAttention A faster alternative to Metal Performance Shaders, a reference implementation of modern GPU algorithms, and a step toward defragmenting the AI ecosystem. Algorithms: Attention Dense (90.5% ALU) Block-Sparse GEMM FP16 (93.3% ALU) FP32 (87.2% ALU) Fused Biases Usage Progamming Language MFA Supports MPSGraph Supports PyTorch Supports CPU C++ (metal-cpp) ✅ ❌ ✅ GPU C++ (Indirect Command Buffers) ✅ ❌ ❌ Swift (iPadOS, Playgrounds) ✅ ✅ ❌ Swift (macOS, Xcode) ✅ ✅ ✅ Predecessor...
Metal FlashAttention A faster alternative to Metal Performance Shaders, a reference implementation of modern GPU algorithms, and a step toward defragmenting the AI ecosystem. Algorithms: Attention Dense (90.5% ALU) Block-Sparse GEMM FP16 (93.3% ALU) FP32 (87.2% ALU) Fused Biases Usage Progamming Language MFA Supports MPSGraph Supports PyTorch Supports CPU C++ (metal-cpp) ✅ ❌ ✅ GPU C++ (Indirect Command Buffers) ✅ ❌ ❌ Swift (iPadOS, Playgrounds) ✅ ✅ ❌ Swift (macOS, Xcode) ✅ ✅ ✅ Predecessor...
| repo | techs | stars | weekly | forks | weekly |
|---|---|---|---|---|---|
| metal-by-example/metal-spatial-rendering | Objective-C++MetalObjective-C | 122 | 0 | 9 | 0 |
| e2b-dev/awesome-ai-agents | 1.8k | +63 | 134 | +9 | |
| CarloLucibello/GraphNeuralNetworks.jl | JuliaJupyter Notebook | 167 | 0 | 41 | 0 |
| italomandara/CXPatcher | ASP.NETSwiftRich Text Format | 522 | +4 | 15 | -1 |
| WeTransfer/WeScan | SwiftOther | 2.7k | 0 | 518 | 0 |
| mosaicml/streaming | PythonOther | 612 | 0 | 83 | 0 |
| canisminor1990/sd-webui-lobe-theme | TypeScriptPythonOther | 1.4k | 0 | 125 | 0 |
| whylabs/langkit | Jupyter NotebookPythonMakefile | 431 | 0 | 38 | 0 |
| VictorKabata/Notflix | KotlinSwiftRuby | 445 | 0 | 62 | 0 |
| olucurious/Awesome-ARKit | Swift | 7.7k | 0 | 924 | 0 |