| easy-ml cargo | 2.3.0 | none | none | Machine learning library providing matrices, named tensors, linear algebra and automatic differentiation aimed at being easy to use | 42K | 2h ago |
| ruSPARSE cargo | 0.1.103 | none | none | Ruda's sparse-computing library, reusing shared tensors, runtimes, and device backends. | 1.1K | 7h ago |
| katakuri cargo | 0.1.0 | none | none | Model ecosystem layer over Zunesha — safetensors weight loading, WordPiece tokenization, architecture specs, and hub access. Never computes. | 24 | 10h ago |
| lattice-inference cargo | 0.13.0 | none | none | Pure Rust transformer inference engine — safetensors loading, SIMD matmul, BGE/Qwen3 embeddings | 54K | 14h ago |
| nuvec-io cargo | 0.0.1 | none | none | Zero-copy safetensors, GGUF and npy loading for nuvec | 16 | 15h ago |
| neura-source cargo | 0.7.4 | none | none | The tensors a store fills from: a named source of payload and quantum bytes | 176 | 18h ago |
| neura-gguf cargo | 0.7.4 | none | none | Reading GGUF files into the tensors a store fills | 156 | 18h ago |
| fuel-formats cargo | 0.20.0 | none | none | Transport-independent parsers for tensor wire formats (safetensors, pickle, GGUF, GGML, imatrix). Operates on impl Read / &[u8] / Cow<[u8]>; knows nothing… | 46 | 18h ago |
| fusion-symbols cargo | 0.4.1 | none | none | Fusion and recoupling symbols for symmetric tensors: exact SU(2) recoupling, and runtime Clebsch-Gordan / F- / R-coefficient generation for SU(N), SO(N), and… | 83 | 18h ago |
| rlx-hub cargo | 0.2.18 | none | none | RLX — HuggingFace shard-aware download + verification: safetensors index parsing, layer→shard planning for distributed loads, resumable fetch, and integrity… | 39 | 22h ago |
| ruda-autodiff cargo | 0.21.58 | none | none | Automatic differentiation graphs, gradients and recomputation for Ruda tensors. | 919 | 1d ago |
| g_math cargo | 0.6.8 | none | none | Multi-domain fixed-point arithmetic with geometric extension: Lie groups, manifolds, ODE solvers, tensors, fiber bundles — pure Rust, zero-float, deterministic | 1.5K | 1d ago |
| magnetar-loader-huggingface cargo | 1.0.0-RC0 | none | none | Hugging Face (safetensors) production model loader for the Magnetar inference runtime. | 11 | 1d ago |
| magnetar-format-safetensors cargo | 1.0.0-RC0 | none | none | safetensors container parser for the Magnetar inference runtime. | 15 | 1d ago |
| baracuda-optim cargo | 0.0.1-alpha.126 | none | none | Optimizer kernels (Adam / LAMB / SGD) for the baracuda CUDA stack, built on the multi_tensor_apply idiom vendored from NVIDIA Apex (BSD-3-Clause). One launch… | 615 | 1d ago |
| rlx-gguf-convert cargo | | – 0.2.18 did not build | – | Convert safetensors / ONNX initializer tensors into GGUF with per-tensor quantization. Useful for shrinking a model's on-disk + in-memory footprint at first… | 282 | 1d ago |
| rlx-tensor cargo | | – 0.2.18 did not build | – | Native ndarray alternative for RLX — NumPy-style tensors that trace, fuse, and run on any backend via the rlx compiler | 1.4K | 2d ago |
| neura-precision cargo | 0.7.4 | none | none | Host packing of tensors into their declared element format | 114 | 2d ago |
| rlx-dduf cargo | 0.2.18 | none | none | RLX — load HuggingFace DDUF (.dduf) ZIP packages of nested safetensors + JSON. | 797 | 2d ago |
| rlx-mlx-io cargo | | – 0.2.18 did not build | – | RLX — load MLX weight layouts (mlx-community safetensors dirs, .npz/.npy) including affine / mxfp dequant. | 1.2K | 2d ago |
| rstsr-cpu-dlpack cargo | 0.10.2 | none | none | DLPack interchange for rstsr CPU tensors (pure Rust, no Python bindings) | 11 | 3d ago |
| proxima-model-interop cargo | 0.1.0 | none | none | Sans-IO GGUF <-> safetensors transform over the format's common denominator (named tensors, dtype, shape, bytes); explicit about what each direction loses. | 11 | 3d ago |
| proxima-safetensors cargo | 0.1.0 | none | none | Sans-IO safetensors reader: header + tensor-directory FSM over chunked bytes, no dequantization | 14 | 3d ago |
| aprender-serve cargo | 0.71.4 | none | none | Pure Rust ML inference engine built from scratch - model serving for GGUF and safetensors | 6.1K | 4d ago |
| edgefirst-codec cargo | 0.35.0 | none | none | Image codec for decoding JPEG/PNG into pre-allocated EdgeFirst tensors | 3.2K | 5d ago |
| turbospark-repack cargo | 0.3.0 | none | none | Safetensors header parsing, ranged-download planning, and int4/int8 repack into the .gturbo packed-expert layout. | 38 | 6d ago |
| aprender-train-inspect cargo | 0.71.1 | none | none | SafeTensors model inspection and format conversion | 516 | 6d ago |
| hf-fetch-model cargo | 0.12.3 | none | none | Download, inspect, and compare HuggingFace models from Rust. Multi-connection parallel downloads plus safetensors, NPZ, GGUF, and PyTorch .pth header… | 4.4K | 6d ago |
| cubecl-utils-rs cargo | 0.2.0 | none | none | Shared CubeCL helpers: GPU tensors, device-limit queries and validated dispatch geometry. | 1.7K | 6d ago |
| tuplet cargo | | – 0.4.0 did not build | – | Tuple-based metric learning: contrastive losses, hard-negative mining, and classical Mahalanobis methods on Burn tensors | 90 | 7d ago |
