IBM / triton-dejavuLinks
Framework to reduce autotune overhead to zero for well known deployments.
☆74Updated 3 weeks ago
Alternatives and similar repositories for triton-dejavu
Users that are interested in triton-dejavu are comparing it to the libraries listed below
Sorting:
- TileFusion is an experimental C++ macro kernel template library that elevates the abstraction level in CUDA C for tile processing.☆88Updated last week
- ☆59Updated last month
- Tritonbench is a collection of PyTorch custom operators with example inputs to measure their performance.☆127Updated this week
- DeeperGEMM: crazy optimized version☆69Updated last month
- ☆70Updated 2 weeks ago
- Quantized Attention on GPU☆44Updated 6 months ago
- ☆80Updated 6 months ago
- ☆96Updated 8 months ago
- GPTQ inference TVM kernel☆40Updated last year
- extensible collectives library in triton☆87Updated 2 months ago
- ☆49Updated 2 weeks ago
- ☆71Updated 2 months ago
- ☆48Updated this week
- A bunch of kernels that might make stuff slower 😉☆46Updated this week
- PyTorch bindings for CUTLASS grouped GEMM.☆93Updated last week
- Transformers components but in Triton☆33Updated 3 weeks ago
- High-speed GEMV kernels, at most 2.7x speedup compared to pytorch baseline.☆109Updated 10 months ago
- Decoding Attention is specially optimized for MHA, MQA, GQA and MLA using CUDA core for the decoding stage of LLM inference.☆36Updated 2 months ago
- A practical way of learning Swizzle☆19Updated 4 months ago
- TritonBench: Benchmarking Large Language Model Capabilities for Generating Triton Operators☆51Updated 3 months ago
- Boosting 4-bit inference kernels with 2:4 Sparsity☆76Updated 9 months ago
- Odysseus: Playground of LLM Sequence Parallelism☆69Updated 11 months ago
- Benchmark code for the "Online normalizer calculation for softmax" paper☆94Updated 6 years ago
- A GPU-optimized system for efficient long-context LLMs decoding with low-bit KV cache.☆36Updated last month
- ☆19Updated 8 months ago
- ☆86Updated 5 months ago
- ☆26Updated last year
- llama INT4 cuda inference with AWQ