CalvinXKY / mfu_calculationLinks
A simple calculation for LLM MFU.
☆38Updated 3 months ago
Alternatives and similar repositories for mfu_calculation
Users that are interested in mfu_calculation are comparing it to the libraries listed below
Sorting:
- Estimate MFU for DeepSeekV3☆24Updated 5 months ago
- Odysseus: Playground of LLM Sequence Parallelism☆70Updated last year
- [NeurIPS 2024] Efficient LLM Scheduling by Learning to Rank☆48Updated 7 months ago
- ☆60Updated last month
- ☆84Updated 3 years ago
- [ICLR 2025] PEARL: Parallel Speculative Decoding with Adaptive Draft Length☆89Updated 2 months ago
- ☆54Updated last year
- Stateful LLM Serving☆73Updated 3 months ago
- ☆77Updated 2 months ago
- ☆86Updated 2 months ago
- ☆74Updated 4 years ago
- Sequence-level 1F1B schedule for LLMs.☆17Updated last year
- ☆49Updated last month
- Compare different hardware platforms via the Roofline Model for LLM inference tasks.☆100Updated last year
- ☆62Updated last year
- Official implementation of ICML 2024 paper "ExCP: Extreme LLM Checkpoint Compression via Weight-Momentum Joint Shrinking".☆48Updated 11 months ago
- Automated Parallelization System and Infrastructure for Multiple Ecosystems☆79Updated 7 months ago
- [USENIX ATC '24] Accelerating the Training of Large Language Models using Efficient Activation Rematerialization and Optimal Hybrid Paral…☆55Updated 10 months ago
- ATC23 AE☆45Updated 2 years ago
- Summary of system papers/frameworks/codes/tools on training or serving large model☆57Updated last year
- PyTorch bindings for CUTLASS grouped GEMM.☆127Updated 5 months ago
- ☆96Updated 9 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 last year
- PyTorch implementation of paper "Response Length Perception and Sequence Scheduling: An LLM-Empowered LLM Inference Pipeline".☆88Updated 2 years ago
- ☆31Updated last month
- GPTQ inference TVM kernel☆40Updated last year
- nnScaler: Compiling DNN models for Parallel Training☆113Updated this week
- Implement Flash Attention using Cute.☆87Updated 6 months ago
- ☆103Updated 7 months ago
- A lightweight design for computation-communication overlap.☆143Updated this week