crates.io "cuda" keyword
vllm-proto 0.6.0 π°
Protobuf messages and gRPC bindings for the vLLM Rust frontend7 versions - Latest release: 8 days ago - 105 thousand downloads total - 95,140 stars on GitHub - 1 maintainer
libceed-sys 0.12.0
Low-level bindings for libCEED library.4 versions - Latest release: almost 3 years ago - 1 dependent package - 2 dependent repositories - 6.61 thousand downloads total - 270 stars on GitHub - 1 maintainer
libceed 0.12.0
High-level interface for libCEED - the efficient high-order discretization library developed by t...4 versions - Latest release: almost 3 years ago - 5.42 thousand downloads total - 270 stars on GitHub - 1 maintainer
h3-rs 0.5.1
MiniMax H3 text/image-to-video with stereo audio in Rust (CUDA or Apple Silicon MLX)6 versions - Latest release: about 9 hours ago - 51 downloads total - 0 stars on GitHub - 1 maintainer
flodl-cli 0.8.0
libtorch manager and GPU diagnostic tool for Rust deep learning9 versions - Latest release: 2 months ago - 280 downloads total - 63 stars on GitHub - 1 maintainer
Top 6.2% on crates.io
18 versions - Latest release: almost 6 years ago - 17 dependent packages - 10 dependent repositories - 123 thousand downloads total - 826 stars on GitHub - 1 maintainer
arrayfire 3.8.0
ArrayFire is a high performance software library for parallel computing with an easy-to-use API. ...18 versions - Latest release: almost 6 years ago - 17 dependent packages - 10 dependent repositories - 123 thousand downloads total - 826 stars on GitHub - 1 maintainer
piper-plus 0.5.2 π°
High-quality neural text-to-speech engine with 8-language support6 versions - Latest release: 17 days ago - 395 downloads total - 238 stars on GitHub - 1 maintainer
piper-plus-cli 0.5.2 π°
CLI tool for Piper Plus neural text-to-speech synthesis6 versions - Latest release: 17 days ago - 153 downloads total - 238 stars on GitHub - 1 maintainer
piper-plus-g2p 0.5.2 π°
Multilingual G2P (Grapheme-to-Phoneme) for TTS β eSpeak-ng free5 versions - Latest release: 17 days ago - 5.65 thousand downloads total - 238 stars on GitHub - 1 maintainer
burn-nn 0.22.1 π°
Neural network building blocks for the Burn deep learning framework22 versions - Latest release: 3 days ago - 956 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
burn-fusion 0.22.1 π°
Kernel fusion backend decorator for the Burn framework36 versions - Latest release: 3 days ago - 3 dependent packages - 1.38 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-cuda 0.22.1 π°
CUDA backend for the Burn framework29 versions - Latest release: 3 days ago - 1.47 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-ir 0.22.1 π°
Intermediate representation for the Burn framework25 versions - Latest release: 3 days ago - 1.5 million downloads total - 16,136 stars on GitHub - 2 maintainers
Top 5.4% on crates.io
49 versions - Latest release: 3 days ago - 9 dependent packages - 19 dependent repositories - 1.96 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-tensor 0.22.1 π°
Tensor library with user-friendly APIs and automatic differentiation support49 versions - Latest release: 3 days ago - 9 dependent packages - 19 dependent repositories - 1.96 million downloads total - 16,136 stars on GitHub - 2 maintainers
Top 8.2% on crates.io
42 versions - Latest release: 3 days ago - 1 dependent package - 15 dependent repositories - 1.21 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-train 0.22.1 π°
Training crate for the Burn framework42 versions - Latest release: 3 days ago - 1 dependent package - 15 dependent repositories - 1.21 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-store 0.22.1 π°
Storage and serialization infrastructure for Burn22 versions - Latest release: 3 days ago - 861 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
Top 7.2% on crates.io
38 versions - Latest release: 6 months ago - 3 dependent packages - 17 dependent repositories - 91.8 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
burn-tensor-testgen 0.21.0 π°
Test generation crate for burn-tensor38 versions - Latest release: 6 months ago - 3 dependent packages - 17 dependent repositories - 91.8 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
burn-rl 0.22.1 π°
RL crate for the Burn framework12 versions - Latest release: 3 days ago - 323 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
Top 5.5% on crates.io
24 versions - Latest release: 12 months ago - 10 dependent packages - 17 dependent repositories - 1.14 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-common 0.19.1 π°
Common crate for the Burn framework24 versions - Latest release: 12 months ago - 10 dependent packages - 17 dependent repositories - 1.14 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-dispatch 0.22.1 π°
Backend dispatch for the Burn framework11 versions - Latest release: 3 days ago - 522 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
pytorch-reader 0.22.1 π°
Reader for PyTorch checkpoint files (.pt/.pth) that depends on neither PyTorch nor Burn3 versions - Latest release: 3 days ago - 2.83 thousand downloads total - 16,136 stars on GitHub - 1 maintainer
burn-vision 0.22.1 π°
Vision processing operations for burn tensors25 versions - Latest release: 3 days ago - 415 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
Top 6.8% on crates.io
46 versions - Latest release: 3 days ago - 2 dependent packages - 19 dependent repositories - 1.9 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-derive 0.22.1 π°
Derive crate for the Burn framework46 versions - Latest release: 3 days ago - 2 dependent packages - 19 dependent repositories - 1.9 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-capture 0.22.1 π°
Graph capture backend for the Burn framework4 versions - Latest release: 3 days ago - 569 downloads total - 16,136 stars on GitHub - 1 maintainer
burn-flex 0.22.1 π°
A fast, portable CPU backend for the Burn framework11 versions - Latest release: 3 days ago - 480 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
burn-einsum 0.22.1 π°
Shared equation parsing and execution planning for Burn's einsum APIs3 versions - Latest release: 3 days ago - 23 thousand downloads total - 16,136 stars on GitHub - 1 maintainer
burn-compute 0.13.2 π°
Compute crate that helps creating high performance async backends.8 versions - Latest release: over 2 years ago - 2 dependent packages - 1 dependent repositories - 176 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
burn-backend 0.22.1 π°
Core backend interfaces and data structures for executing tensor operations in Burn.17 versions - Latest release: 3 days ago - 813 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
burn-hip 0.16.1 π°
ROCm HIP backend for the Burn framework3 versions - Latest release: over 1 year ago - 154 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
burn-tch 0.22.1 π°
[Deprecated] LibTorch backend for the Burn framework - use burn-cuda, burn-rocm, burn-wgpu, burn-...44 versions - Latest release: 3 days ago - 3 dependent packages - 17 dependent repositories - 983 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
burn-jit 0.16.1 π°
Generic backend that can be compiled just-in-time to any shader language target7 versions - Latest release: over 1 year ago - 1 dependent package - 280 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
Top 7.1% on crates.io
43 versions - Latest release: 3 days ago - 2 dependent packages - 17 dependent repositories - 1.9 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-core 0.22.1 π°
Flexible and Comprehensive Deep Learning Framework in Rust43 versions - Latest release: 3 days ago - 2 dependent packages - 17 dependent repositories - 1.9 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-communication 0.22.1 π°
Abstractions for network communication for Burn22 versions - Latest release: 3 days ago - 399 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
Top 7.1% on crates.io
49 versions - Latest release: 3 days ago - 7 dependent packages - 17 dependent repositories - 1.88 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn 0.22.1 π°
Flexible and Comprehensive Deep Learning Framework in Rust49 versions - Latest release: 3 days ago - 7 dependent packages - 17 dependent repositories - 1.88 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-cpu 0.22.1 π°
LLVM-based CPU backend for the Burn framework22 versions - Latest release: 3 days ago - 717 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
burn-linalg 0.22.1 π°
Linear algebra operations for Burn tensors3 versions - Latest release: 3 days ago - 22.6 thousand downloads total - 16,136 stars on GitHub - 1 maintainer
burn-pack 0.22.1 π°
The burnpack binary serialization format for Burn6 versions - Latest release: 3 days ago - 32.3 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
Top 7.2% on crates.io
46 versions - Latest release: 3 days ago - 2 dependent packages - 19 dependent repositories - 1.23 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-dataset 0.22.1 π°
Library with simple dataset APIs for creating ML data pipelines46 versions - Latest release: 3 days ago - 2 dependent packages - 19 dependent repositories - 1.23 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-remote 0.22.1 π°
Backend router decorator over the network.27 versions - Latest release: 3 days ago - 514 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
burn-router 0.22.1 π°
Multi-backend router decorator for the Burn framework27 versions - Latest release: 3 days ago - 1.36 million downloads total - 16,136 stars on GitHub - 2 maintainers
Top 6.2% on crates.io
44 versions - Latest release: 3 days ago - 8 dependent packages - 17 dependent repositories - 1.86 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-ndarray 0.22.1 π°
[Deprecated] Ndarray backend for the Burn framework - use burn-flex, burn-cuda, burn-rocm, or bur...44 versions - Latest release: 3 days ago - 8 dependent packages - 17 dependent repositories - 1.86 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-backend-extension 0.22.1 π°
Burn backend extension proc macro6 versions - Latest release: 3 days ago - 32.4 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
burn-wgpu 0.22.1 π°
WGPU backend for the Burn framework39 versions - Latest release: 3 days ago - 3 dependent packages - 2 dependent repositories - 1.72 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-collective 0.21.0 π°
Backend extension for collective calculations.16 versions - Latest release: 6 months ago - 398 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
Top 6.0% on crates.io
44 versions - Latest release: 3 days ago - 9 dependent packages - 17 dependent repositories - 1.78 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-autodiff 0.22.1 π°
Automatic differentiation backend for the Burn framework44 versions - Latest release: 3 days ago - 9 dependent packages - 17 dependent repositories - 1.78 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-rocm 0.22.1 π°
ROCm HIP backend for the Burn framework25 versions - Latest release: 3 days ago - 1.27 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-optim 0.22.1 π°
Optimizer building blocks for the Burn deep learning framework22 versions - Latest release: 3 days ago - 950 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
burn-cubecl-fusion 0.22.1 π°
Provide optimizations that can be used with cubecl based backends.25 versions - Latest release: 3 days ago - 1.1 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-cubecl 0.22.1 π°
Generic backend that can be compiled just-in-time to any shader language target25 versions - Latest release: 3 days ago - 1.36 million downloads total - 16,136 stars on GitHub - 2 maintainers
burn-signal 0.22.1 π°
Signal processing operations for Burn tensors3 versions - Latest release: 3 days ago - 22.5 thousand downloads total - 16,136 stars on GitHub - 1 maintainer
burn-std 0.22.1 π°
Core types and utilities shared across the Burn ecosystem.17 versions - Latest release: 3 days ago - 814 thousand downloads total - 16,136 stars on GitHub - 2 maintainers
burn-candle 0.21.0 π°
[Deprecated] Candle backend for the Burn framework - use burn-cubecl, burn-ndarray, or burn-tch i...33 versions - Latest release: 6 months ago - 1 dependent package - 5 dependent repositories - 1.58 million downloads total - 16,136 stars on GitHub - 2 maintainers
with-gpu 0.6.0
Intelligent GPU selection wrapper for CUDA commands4 versions - Latest release: 2 months ago - 122 downloads total - 0 stars on GitHub - 1 maintainer
baracuda-nvshmem 0.0.1-alpha.147
Safe Rust wrappers for NVIDIA NVSHMEM host API (symmetric-heap one-sided GPU communication).34 versions - Latest release: about 23 hours ago - 626 downloads total - 6 stars on GitHub - 1 maintainer
baracuda-kernels-types 0.0.1-alpha.147
Shared type vocabulary for the baracuda ML kernel facade: Element / IntElement / FpElement / Bias...70 versions - Latest release: about 24 hours ago - 22.9 thousand downloads total - 6 stars on GitHub - 1 maintainer
baracuda-cupti-sys 0.0.1-alpha.147
Raw FFI bindings and dynamic loader for CUPTI (CUDA Profiling Tools Interface).77 versions - Latest release: about 24 hours ago - 1.64 thousand downloads total - 6 stars on GitHub - 1 maintainer
baracuda-cutensor-sys 0.0.1-alpha.147
Raw FFI bindings and dynamic loader for NVIDIA cuTENSOR (tensor contraction).77 versions - Latest release: about 24 hours ago - 1.67 thousand downloads total - 6 stars on GitHub - 1 maintainer
mwa_hyperdrive 0.8.1
Calibration software for the Murchison Widefield Array (MWA) radio telescope10 versions - Latest release: about 1 month ago - 6.73 thousand downloads total - 137 stars on GitHub - 2 maintainers
coglet 0.23.1
High-performance prediction server for Cog ML models27 versions - Latest release: 3 days ago - 537 downloads total - 9,490 stars on GitHub - 1 maintainer
media-pp 0.4.1
A GStreamer-flavored media pipeline library built on FFmpeg. Decode, capture, composite, encode, ...15 versions - Latest release: 7 days ago - 261 downloads total - 3 stars on GitHub - 1 maintainer
tree-sitter-cuda 0.21.2
CUDA grammar for tree-sitter14 versions - Latest release: about 1 month ago - 1 dependent package - 1 dependent repositories - 86.3 thousand downloads total - 36 stars on GitHub - 1 maintainer
candle-cuda-vmm 0.1.1
CUDA Virtual Memory Management bindings for elastic KV cache allocation in Candle2 versions - Latest release: 11 months ago - 193 downloads total - 2 stars on GitHub - 1 maintainer
sumavia 0.0.0
Rust LLM inference stack for MoE models on consumer GPUs (in development).1 version - Latest release: 1 day ago - 0 downloads total - 1 maintainer
cubecl 0.11.2
Multi-platform high-performance compute language extension for Rust.29 versions - Latest release: 1 day ago - 1.66 million downloads total - 2,439 stars on GitHub - 2 maintainers
cubecl-server 0.11.2
Toolkit for implementing a CubeCL runtime: memory pools, streams, drivers and the compilation pip...4 versions - Latest release: 1 day ago - 29.2 thousand downloads total - 2,439 stars on GitHub - 1 maintainer
cubecl-core 0.11.2
CubeCL core create29 versions - Latest release: 1 day ago - 1.67 million downloads total - 2,439 stars on GitHub - 1 maintainer
basal-client 0.7.0
Async Rust client for the native Basal HTTP API7 versions - Latest release: 2 days ago - 66 downloads total - 9 stars on GitHub - 1 maintainer
hanzo-kernel 0.2.43 π°
Hanzo's first-party GPU kernel DSL: one Rust source, lowered to CUDA/ROCm/Vulkan/Metal.25 versions - Latest release: 16 days ago - 651 downloads total - 2 stars on GitHub - 1 maintainer
hanzo-flash-attn-build 0.11.10 π°
Build-time helper that fetches and resolves the pinned NVIDIA CUTLASS include directory for the h...8 versions - Latest release: 16 days ago - 1.1 thousand downloads total - 2 stars on GitHub - 1 maintainer
xlog-ir 0.11.0 π°
Relational intermediate representation for XLOG query plans4 versions - Latest release: 3 months ago - 851 downloads total - 6 stars on GitHub - 1 maintainer
xlog-gpu 0.11.0 π°
High-level Rust API for running XLOG programs on NVIDIA GPUs4 versions - Latest release: 3 months ago - 456 downloads total - 6 stars on GitHub - 1 maintainer
xlog-stats 0.11.0 π°
Runtime statistics and optimizer feedback utilities for XLOG4 versions - Latest release: 3 months ago - 846 downloads total - 6 stars on GitHub - 1 maintainer
xlog-cli 0.11.0 π°
Command-line interface for deterministic and probabilistic XLOG execution4 versions - Latest release: 3 months ago - 356 downloads total - 6 stars on GitHub - 1 maintainer
xlog-runtime 0.11.0 π°
Runtime executor and relation store for XLOG4 versions - Latest release: 3 months ago - 719 downloads total - 6 stars on GitHub - 1 maintainer
xlog-prob 0.11.0 π°
Probabilistic inference engines for XLOG4 versions - Latest release: 3 months ago - 459 downloads total - 6 stars on GitHub - 1 maintainer
xlog-logic 0.11.0 π°
Parser, compiler, and optimizer for XLOG logic programs4 versions - Latest release: 3 months ago - 679 downloads total - 6 stars on GitHub - 1 maintainer
xlog-core 0.11.0 π°
Core types, traits, and error surfaces shared across XLOG4 versions - Latest release: 3 months ago - 1.07 thousand downloads total - 6 stars on GitHub - 1 maintainer
xlog-solve 0.11.0 π°
Solver services used by XLOG exact inference and verification layers4 versions - Latest release: 3 months ago - 559 downloads total - 6 stars on GitHub - 1 maintainer
xlog-cuda 0.11.0 π°
CUDA kernel provider, buffers, and interop for XLOG4 versions - Latest release: 3 months ago - 815 downloads total - 6 stars on GitHub - 1 maintainer
ptxlint 0.0.4
Static analysis and lints for NVIDIA PTX, no GPU needed. δΈιθ¦ζΎε‘η NVIDIA PTX ιζεζε·₯ε ·4 versions - Latest release: about 1 month ago - 54 downloads total - 0 stars on GitHub - 1 maintainer
bylazora-core 0.5.1
Byte-exact COBOL mainframe migration engine: one binary runs the equivalence gate and the product...3 versions - Latest release: about 1 month ago - 29 downloads total - 0 stars on GitHub - 1 maintainer
crush-ptx 0.7.1
CASM -> PTX (NVIDIA GPU assembly) emission backend for Crush kernel programs3 versions - Latest release: 3 days ago - 31 downloads total - 0 stars on GitHub - 1 maintainer
tritium-core 1.1.0-rc.2
Foundational ternary types, dtypes, scaling schemes, and reference math for Tritium.2 versions - Latest release: about 2 months ago - 1.86 thousand downloads total - 28 stars on GitHub - 1 maintainer
tritium-build-info 1.1.0-rc.2
Build-script support for source-bound Tritium release evidence.2 versions - Latest release: about 2 months ago - 1.63 thousand downloads total - 28 stars on GitHub - 1 maintainer
tritium-mcu 1.1.0-rc.2
no_std fixed-arena executor for the ternary codec on a $2 MCU (Cortex-M55/M33), backed by the bit...2 versions - Latest release: about 2 months ago - 682 downloads total - 28 stars on GitHub - 1 maintainer
tritium-burn 1.1.0-rc.2
burn interop for Tritium: a host round-trip op that runs Tritium's ternary (BitNet b1.58) mpGEMM ...2 versions - Latest release: about 2 months ago - 718 downloads total - 28 stars on GitHub - 1 maintainer
tritium-onnx 1.1.0-rc.2
ONNX Runtime interop for Tritium: packed ternary mpGEMM and selected-row embedding kernels plus c...2 versions - Latest release: about 2 months ago - 656 downloads total - 28 stars on GitHub - 1 maintainer
tritium-candle 1.1.0-rc.2
candle interop for Tritium: a `CustomOp1` that runs Tritium's ternary (BitNet b1.58) mpGEMM on a ...2 versions - Latest release: about 2 months ago - 719 downloads total - 28 stars on GitHub - 1 maintainer
tritium-metal 1.1.0-rc.2
Apple-Silicon execution backend: metal-rs host side + an add/sub/skip ternary mpGEMM compute kern...2 versions - Latest release: about 2 months ago - 690 downloads total - 28 stars on GitHub - 1 maintainer
tritium-rocm 1.1.0-rc.2
AMD ROCm/HIP execution backend: raw HIP-runtime FFI host side + an addition-only ternary mpGEMM k...2 versions - Latest release: about 2 months ago - 644 downloads total - 28 stars on GitHub - 1 maintainer
tritium-train 1.1.0-rc.2
STE autograd, QAT, optimizer, and LoRA for ternary BitNet models (single-node).2 versions - Latest release: about 2 months ago - 1.49 thousand downloads total - 28 stars on GitHub - 1 maintainer
tritium-wasm 1.1.0-rc.2
Scalar WebAssembly execution backend for wasm32-wasip1: the reference ternary mpGEMM, no host-onl...2 versions - Latest release: about 2 months ago - 600 downloads total - 28 stars on GitHub - 1 maintainer
tritium-ffi 1.1.0-rc.2
C ABI (cdylib + staticlib) for Tritium ternary-model inference: load a GGUF model and greedily ge...2 versions - Latest release: about 2 months ago - 701 downloads total - 28 stars on GitHub - 1 maintainer
tritium-runtime 1.1.0-rc.2
Backend registry and dispatch β backends self-register via linkme; no concrete backend deps.2 versions - Latest release: about 2 months ago - 1.59 thousand downloads total - 28 stars on GitHub - 1 maintainer
tritium-salt 1.1.0-rc.2
High-level, resumable orchestration for SALT V2 additive ternarization.2 versions - Latest release: about 2 months ago - 644 downloads total - 28 stars on GitHub - 1 maintainer
tritium-nn 1.1.0-rc.2
Inference layer over the ternary backend: nn ops, KV cache, and the BitNet model runner.2 versions - Latest release: about 2 months ago - 1.46 thousand downloads total - 28 stars on GitHub - 1 maintainer
tritium-cpu 1.1.0-rc.2
CPU execution backend: AVX2 (and scalar fallback) ternary mpGEMM with runtime dispatch.2 versions - Latest release: about 2 months ago - 1.52 thousand downloads total - 28 stars on GitHub - 1 maintainer
tritium-cli 1.1.0-rc.2
The `tritium` command-line tool: inspect GGUF files and list registered backends.2 versions - Latest release: about 2 months ago - 29 downloads total - 28 stars on GitHub - 1 maintainer
tritium-format 1.1.0-rc.2
Ternary weight packing (TQ1_0/TQ2_0) and GGUF I/O β the single source of truth for on-disk layout.2 versions - Latest release: about 2 months ago - 1.76 thousand downloads total - 28 stars on GitHub - 1 maintainer
glazier-core 0.3.0
Cellular Potts lattice, energies and serial sweep7 versions - Latest release: about 1 month ago - 120 downloads total - 0 stars on GitHub - 1 maintainer
Related Keywords
rust
795
gpu
700
machine-learning
314
metal
274
tensor
272
deep-learning
247
neural-network
245
nvidia
226
inference
197
ml
177
compiler
173
llm
163
mlx
151
rust-lang
134
gpu-computing
115
computer-vision
114
gpgpu
109
linalg
106
autodiff
102
webgpu
96
pytorch
90
cuda-programming
89
cpu
86
transformers
85
rocm
83
onnx
82
vulkan
82
text-to-speech
82
tts
82
ir
79
pure-rust
78
speech-recognition
78
asr
78
ai
76
vlm
76
nvidia-gpu
74
npu
69
fpga
68
cortexm
68
ffi
63
compute
60
runtime
59
wasm
56
wgpu
56
driver
54
infra
53
inference-engine
52
hpc
51
cross-platform
50
scientific-computing
50
quantization
48
ndarray
48
autograd
47
parallel-computing
47
kernel-fusion
44
jit
41
artificial-intelligence
39
llm-inference
38
actor
37
kernel
37
edge-ai
37
gguf
36
gpu-acceleration
36
opencl
35
object-detection
34
cli
34
ternary
34
image-processing
32
ptx
31
transformer
31
cudnn
31
spir-v
30
cublas
30
codec
30
ht-j2k
30
htj2k
30
jpeg2000
30
transcoding
30
jpeg2k
30
jpeg-to-jpeg2000
30
jpeg-2000
30
j2k
30
image-codec
30
image-compression
30
jpeg
30
vyre
29
gpu-programming
28
python
28
ai-agents
28
simd
27
video
27
nlp
25
openai-api
24
bitnet
23
distributed-training
23
calibration
21
gpu-native-actors
21
blas
21
npp
21
data
21