Edward-Sun / gpt-acceleraLinks
Simple and efficient pytorch-native transformer training and inference (batched)
☆79Updated last year
Alternatives and similar repositories for gpt-accelera
Users that are interested in gpt-accelera are comparing it to the libraries listed below
Sorting:
- Code and Configs for Asynchronous RLHF: Faster and More Efficient RL for Language Models☆64Updated 5 months ago
- Language models scale reliably with over-training and on downstream tasks☆100Updated last year
- ☆53Updated last year
- ☆49Updated last year
- Using FlexAttention to compute attention with different masking patterns☆44Updated last year
- A toolkit for scaling law research ⚖☆52Updated 8 months ago
- ☆56Updated last year
- Official repository of paper "RNNs Are Not Transformers (Yet): The Key Bottleneck on In-context Retrieval"☆27Updated last year
- Awesome Triton Resources☆34Updated 5 months ago
- Code for NeurIPS 2024 Spotlight: "Scaling Laws and Compute-Optimal Training Beyond Fixed Training Durations"☆84Updated 11 months ago
- The simplest implementation of recent Sparse Attention patterns for efficient LLM inference.☆89Updated 2 months ago
- A scalable asynchronous reinforcement learning implementation with in-flight weight updates.☆145Updated last week
- The source code of our work "Prepacking: A Simple Method for Fast Prefilling and Increased Throughput in Large Language Models" [AISTATS …☆60Updated 11 months ago
- Long Context Extension and Generalization in LLMs☆60Updated last year
- ☆33Updated last year
- ☆104Updated last year
- Understand and test language model architectures on synthetic tasks.☆229Updated last week
- ☆54Updated 3 months ago
- NAACL '24 (Best Demo Paper RunnerUp) / MlSys @ NeurIPS '23 - RedCoast: A Lightweight Tool to Automate Distributed Training and Inference☆68Updated 9 months ago
- Replicating O1 inference-time scaling laws☆90Updated 10 months ago
- ☆101Updated last year
- A fusion of a linear layer and a cross entropy loss, written for pytorch in triton.☆70Updated last year
- Linear Attention Sequence Parallelism (LASP)☆86Updated last year
- ☆85Updated last year
- CUDA and Triton implementations of Flash Attention with SoftmaxN.☆73Updated last year
- Fast and memory-efficient exact attention☆70Updated 7 months ago
- ☆83Updated last year
- ☆38Updated last year
- Stick-breaking attention☆60Updated 3 months ago
- Triton Implementation of HyperAttention Algorithm☆48Updated last year