Alternatives and similar repositories for moe

Users that are interested in moe are comparing it to the libraries listed below

• amirzandieh / HyperAttention
  Triton Implementation of HyperAttention Algorithm
  ☆48Updated last year
• RobertCsordas / moe_attention
  Official repository for the paper "SwitchHead: Accelerating Transformers with Mixture-of-Experts Attention"
  ☆98Updated 9 months ago
• TRI-ML / linear_open_lm
  A repository for research on medium sized language models.
  ☆77Updated last year
• RobertCsordas / moe_layer
  sigma-MoE layer
  ☆20Updated last year
• recursal / GoldFinch-paper
  GoldFinch and other hybrid transformer components
  ☆45Updated 11 months ago
• RobertCsordas / moeut
  ☆82Updated 10 months ago
• google-deepmind / asyncdiloco
  ☆45Updated last year
• shreyansh26 / Attention-Mask-Patterns
  Using FlexAttention to compute attention with different masking patterns
  ☆44Updated 9 months ago
• kaiokendev / cutoff-len-is-context-len
  Demonstration that finetuning RoPE model on larger sequences than the pre-trained model adapts the model context limit
  ☆63Updated 2 years ago
• sustcsonglin / mamba-triton
  ☆48Updated last year
• dangxingyu / rnn-icrag
  Official repository of paper "RNNs Are Not Transformers (Yet): The Key Bottleneck on In-context Retrieval"
  ☆27Updated last year
• lucidrains / transformer-lm-gan
  Explorations into adversarial losses on top of autoregressive loss for language modeling
  ☆37Updated 4 months ago
• smonsays / hypernetwork-attention
  Official code for the paper "Attention as a Hypernetwork"
  ☆39Updated last year
• renll / SeqBoat
  [NeurIPS 2023] Sparse Modular Activation for Efficient Sequence Modeling
  ☆37Updated last year
• HazyResearch / prefix-linear-attention
  ☆55Updated last year
• glassroom / heinsen_attention
  Reference implementation of "Softmax Attention with Constant Cost per Token" (Heinsen, 2024)
  ☆24Updated last year
• epfml / DenseFormer
  ☆81Updated last year
• BlinkDL / LinearAttentionArena
  Here we will test various linear attention designs.
  ☆60Updated last year
• kyegomez / Blockwise-Parallel-Transformer
  32 times longer context window than vanilla Transformers and up to 4 times longer than memory efficient Transformers.
  ☆48Updated 2 years ago
• OpenNLPLab / HGRN2
  HGRN2: Gated Linear RNNs with State Expansion
  ☆55Updated 10 months ago
• Zyphra / Zyda_processing
  ☆35Updated last year
• proger / hippogriff
  Griffin MQA + Hawk Linear RNN Hybrid
  ☆87Updated last year
• lucidrains / pause-transformer
  Yet another random morning idea to be quickly tried and architecture shared if it works; to allow the transformer to pause for any amount…
  ☆54Updated last year
• ctlllll / understanding_llm_benchmarks
  Understanding the correlation between different LLM benchmarks
  ☆29Updated last year
• sjelassi / transformers_ssm_copy
  ☆32Updated last year
• eth-easl / fmengine
  Utilities for Training Very Large Models
  ☆58Updated 9 months ago
• nanowell / Q-Sparse-LLM
  My Implementation of Q-Sparse: All Large Language Models can be Fully Sparsely-Activated
  ☆32Updated 10 months ago
• lucidrains / gateloop-transformer
  Implementation of GateLoop Transformer in Pytorch and Jax
  ☆89Updated last year
• IST-DASLab / SparseFinetuning
  Repository for Sparse Finetuning of LLMs via modified version of the MosaicML llmfoundry
  ☆42Updated last year
• yikangshen / megablocks
  ☆20Updated last year