m-a-n-i-f-e-s-t / power-attentionLinks
Attention Kernels for Symmetric Power Transformers
☆115Updated last week
Alternatives and similar repositories for power-attention
Users that are interested in power-attention are comparing it to the libraries listed below
Sorting:
- DeMo: Decoupled Momentum Optimization☆190Updated 9 months ago
- H-Net Dynamic Hierarchical Architecture☆80Updated last month
- ☆102Updated last month
- seqax = sequence modeling + JAX☆166Updated last month
- supporting pytorch FSDP for optimizers☆84Updated 8 months ago
- A MAD laboratory to improve AI architecture designs 🧪☆127Updated 8 months ago
- The simplest, fastest repository for training/finetuning medium-sized GPTs.☆156Updated 2 months ago
- ☆53Updated last year
- Minimal (400 LOC) implementation Maximum (multi-node, FSDP) GPT training☆131Updated last year
- Accelerated First Order Parallel Associative Scan☆187Updated last year
- Normalized Transformer (nGPT)☆187Updated 9 months ago
- Tree Attention: Topology-aware Decoding for Long-Context Attention on GPU clusters☆129Updated 9 months ago
- Understand and test language model architectures on synthetic tasks.☆222Updated last month
- ☆82Updated last year
- ☆53Updated last year
- 🧱 Modula software package☆231Updated 2 weeks ago
- A JAX-native LLM Post-Training Library☆128Updated last week
- an open source reproduction of NVIDIA's nGPT (Normalized Transformer with Representation Learning on the Hypersphere)☆105Updated 5 months ago
- 📄Small Batch Size Training for Language Models☆57Updated last week
- NanoGPT-speedrunning for the poor T4 enjoyers☆70Updated 4 months ago
- Experiment of using Tangent to autodiff triton☆80Updated last year
- ☆57Updated 11 months ago
- ☆87Updated last year
- Large scale 4D parallelism pre-training for 🤗 transformers in Mixture of Experts *(still work in progress)*☆87Updated last year
- ☆65Updated 9 months ago
- ☆40Updated 4 months ago
- Dion optimizer algorithm☆334Updated this week
- Custom triton kernels for training Karpathy's nanoGPT.☆19Updated 10 months ago
- Pytorch implementation of the PEER block from the paper, Mixture of A Million Experts, by Xu Owen He at Deepmind☆127Updated last year
- ☆55Updated last year