epfml / DenseFormer
☆79Updated 11 months ago
Alternatives and similar repositories for DenseFormer:
Users that are interested in DenseFormer are comparing it to the libraries listed below
- ☆53Updated last year
- some common Huggingface transformers in maximal update parametrization (µP)☆80Updated 3 years ago
- supporting pytorch FSDP for optimizers☆79Updated 3 months ago
- Implementation of GateLoop Transformer in Pytorch and Jax☆87Updated 9 months ago
- The simplest, fastest repository for training/finetuning medium-sized GPTs.☆100Updated 4 months ago
- Explorations into the proposal from the paper "Grokfast, Accelerated Grokking by Amplifying Slow Gradients"☆98Updated 2 months ago
- A MAD laboratory to improve AI architecture designs 🧪☆107Updated 3 months ago
- Collection of autoregressive model implementation☆83Updated last month
- A repository for research on medium sized language models.☆77Updated 9 months ago
- Minimal (400 LOC) implementation Maximum (multi-node, FSDP) GPT training☆122Updated 11 months ago
- Implementation of Infini-Transformer in Pytorch☆109Updated 2 months ago
- ☆49Updated last year
- A fast implementation of T5/UL2 in PyTorch using Flash Attention☆85Updated last week
- Official repository for the paper "Approximating Two-Layer Feedforward Networks for Efficient Transformers"☆36Updated last year
- Yet another random morning idea to be quickly tried and architecture shared if it works; to allow the transformer to pause for any amount…☆53Updated last year
- Understand and test language model architectures on synthetic tasks.☆184Updated last week
- Experiments for efforts to train a new and improved t5☆77Updated 11 months ago
- Triton Implementation of HyperAttention Algorithm☆47Updated last year
- ☆89Updated last month
- ☆75Updated 8 months ago
- ☆51Updated 9 months ago
- GoldFinch and other hybrid transformer components☆45Updated 7 months ago
- Explorations into the recently proposed Taylor Series Linear Attention☆94Updated 7 months ago
- Exploration into the proposed "Self Reasoning Tokens" by Felipe Bonetto☆55Updated 10 months ago
- ☆43Updated last year
- Train a SmolLM-style llm on fineweb-edu in JAX/Flax with an assortment of optimizers.☆17Updated last week
- Code for NeurIPS 2024 Spotlight: "Scaling Laws and Compute-Optimal Training Beyond Fixed Training Durations"☆71Updated 4 months ago
- ☆73Updated 7 months ago
- Large scale 4D parallelism pre-training for 🤗 transformers in Mixture of Experts *(still work in progress)*☆81Updated last year