tile-ai / TileAttentionLinks
☆32Updated this week
Alternatives and similar repositories for TileAttention
Users that are interested in TileAttention are comparing it to the libraries listed below
Sorting:
- ☆49Updated 2 weeks ago
- Accelerate LLM preference tuning via prefix sharing with a single line of code☆41Updated last month
- ☆22Updated last year
- Benchmark tests supporting the TiledCUDA library.☆16Updated 6 months ago
- FlexAttention w/ FlashAttention3 Support☆26Updated 8 months ago
- Quantized Attention on GPU☆44Updated 6 months ago
- Best practices for testing advanced Mixtral, DeepSeek, and Qwen series MoE models using Megatron Core MoE.☆17Updated this week
- ☆20Updated last month
- ☆71Updated 2 weeks ago
- triton ver of gqa flash attn, based on the tutorial☆11Updated 10 months ago
- Open deep learning compiler stack for cpu, gpu and specialized accelerators☆18Updated last week
- Odysseus: Playground of LLM Sequence Parallelism☆70Updated 11 months ago
- The official implementation of paper: SimLayerKV: A Simple Framework for Layer-Level KV Cache Reduction.☆46Updated 7 months ago
- Transformers components but in Triton☆33Updated 3 weeks ago
- [ACL 2024] RelayAttention for Efficient Large Language Model Serving with Long System Prompts☆39Updated last year
- Official implementation of "The Sparse Frontier: Sparse Attention Trade-offs in Transformer LLMs"☆32Updated last month
- ☆31Updated last year
- ☆20Updated 2 months ago
- 方便扩展的Cuda算子理解和优化框架,仅用在学习使用☆15Updated 11 months ago
- ☆19Updated 8 months ago
- Beyond KV Caching: Shared Attention for Efficient LLMs☆19Updated 10 months ago
- Implementation of IceFormer: Accelerated Inference with Long-Sequence Transformers on CPUs (ICLR 2024).☆25Updated 11 months ago
- ☆14Updated 2 years ago
- GPTQ inference TVM kernel☆40Updated last year
- Dynamic Context Selection for Efficient Long-Context LLMs☆28Updated 2 weeks ago
- Framework to reduce autotune overhead to zero for well known deployments.☆74Updated 3 weeks ago
- Vocabulary Parallelism☆19Updated 2 months ago
- ☆49Updated last year
- ☆59Updated last month
- DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling☆15Updated 3 weeks ago