Back to search
Top Contributors
88
commits
Pure Gleam tensor library with quantization (INT8, NF4, AWQ), Flash Attention, and 2:4 Sparsity - 7.5x memory multiplication
Stars
2
Forks
0
Watchers
2
Open Issues
0
Overall repository health assessment
No package.json found
This might not be a Node.js project
88
commits
release: v2.1.0 — validated Ubuntu build, VIVA mycelium compatibility
13bbb2eView on GitHubfix: move bench/examples under viva_tensor namespace for Hex publishing
d7fdb1cView on GitHubrelease: v2.0.0 — modular NIF, CUDA Tensor Cores, 2:4 sparsity, 1854 TOPS peak
ebb7459View on GitHubtest(bench): add comprehensive benchmark suite for all backends
a946e0bView on GitHubdocs: update README and CHANGELOG with multi-backend architecture and benchmarks
752567aView on GitHubfeat(gleam): add FFI bindings for LNS, Horde, HDC, CUDA, sparse and quantization
238fb17View on GitHubfeat(erlang): expand NIF wrapper with all new backends and operations
72bcaaaView on GitHubbuild: update build system for multi-backend compilation
a3a93cfView on GitHubfeat(cuda): add CudaTensor, FP16/INT8 Tensor Cores, async pipeline, sparsity and SageAttention
087ebe4View on GitHubfeat(nif): expand NIF layer with BLAS, CUDA, LNS, Horde, HDC and quantized backends
7ef87ebView on GitHubfeat(zig): add LNS, Horde physics, HDC and quantized matmul SIMD kernels
5b77d83View on GitHubfeat: Introduce extensive benchmarking for CUDA-accelerated tensor operations and refactor native Zig NIF integration.
a16a775View on GitHubfeat: Implement Apple Accelerate BLAS/vDSP bindings for hardware-accelerated tensor operations and add comprehensive benchmarks.
465ed0cView on GitHubfeat: Overhaul benchmarking system, add INT8/NF4 quantization and FP32 CUDA GEMM support.
a4d20b0View on GitHubfeat: Implement CUDA/cuBLAS backend for DGEMM and SGEMM in viva_tensor using dynamic library loading.
31b460aView on GitHub