edwardmilsom / function-space-learning-rates-paperLinks
Code for the paper "Function-Space Learning Rates"
☆23Updated 2 months ago
Alternatives and similar repositories for function-space-learning-rates-paper
Users that are interested in function-space-learning-rates-paper are comparing it to the libraries listed below
Sorting:
- ☆19Updated 3 months ago
- ☆34Updated 11 months ago
- 📄Small Batch Size Training for Language Models☆42Updated 2 weeks ago
- ☆56Updated 10 months ago
- JAX Scalify: end-to-end scaled arithmetics☆16Updated 9 months ago
- Code for NeurIPS 2024 Spotlight: "Scaling Laws and Compute-Optimal Training Beyond Fixed Training Durations"☆81Updated 9 months ago
- ☆32Updated 10 months ago
- ☆87Updated last year
- ☆11Updated 5 months ago
- ☆53Updated last year
- Triton Implementation of HyperAttention Algorithm☆48Updated last year
- Unofficial but Efficient Implementation of "Mamba: Linear-Time Sequence Modeling with Selective State Spaces" in JAX☆86Updated last year
- Engineering the state of RNN language models (Mamba, RWKV, etc.)☆32Updated last year
- Implementation of Gradient Agreement Filtering, from Chaubard et al. of Stanford, but for single machine microbatches, in Pytorch☆25Updated 7 months ago
- Minimal (400 LOC) implementation Maximum (multi-node, FSDP) GPT training☆130Updated last year
- Transformer with Mu-Parameterization, implemented in Jax/Flax. Supports FSDP on TPU pods.☆32Updated 2 months ago
- ☆33Updated last year
- A basic pure pytorch implementation of flash attention☆16Updated 9 months ago
- ☆82Updated last year
- Fork of Flame repo for training of some new stuff in development☆15Updated last month
- Minimal (truly) muP implementation, consistent with TP4 and TP5 papers notation☆14Updated 3 months ago
- 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
- ☆85Updated last year
- ☆31Updated last year
- Automatically take good care of your preemptible TPUs☆36Updated 2 years ago
- Official code for the paper "Attention as a Hypernetwork"☆40Updated last year
- An implementation of the Llama architecture, to instruct and delight☆21Updated 2 months ago
- ☆23Updated 8 months ago
- Deep Networks Grok All the Time and Here is Why☆37Updated last year
- supporting pytorch FSDP for optimizers☆84Updated 8 months ago