arogozhnikov / adamw_bfloat16Links
AdamW optimizer for bfloat16 models in pytorch 🔥.
☆32Updated 11 months ago
Alternatives and similar repositories for adamw_bfloat16
Users that are interested in adamw_bfloat16 are comparing it to the libraries listed below
Sorting:
- Engineering the state of RNN language models (Mamba, RWKV, etc.)☆32Updated last year
- AdaCat☆49Updated 2 years ago
- ☆32Updated last year
- A place to store reusable transformer components of my own creation or found on the interwebs☆56Updated 3 weeks ago
- Standalone pre-training recipe with JAX+Flax☆32Updated 2 years ago
- Triton Implementation of HyperAttention Algorithm☆48Updated last year
- ☆53Updated last year
- Implementation of Token Shift GPT - An autoregressive model that solely relies on shifting the sequence space for mixing☆50Updated 3 years ago
- The official Languini Kitchen repository☆14Updated last year
- My explorations into editing the knowledge and memories of an attention network☆35Updated 2 years ago
- LayerNorm(SmallInit(Embedding)) in a Transformer to improve convergence☆60Updated 3 years ago
- Tensor Parallelism with JAX + Shard Map☆11Updated last year
- [NeurIPS 2023] Sparse Modular Activation for Efficient Sequence Modeling☆36Updated last year
- Parallel Associative Scan for Language Models☆18Updated last year
- ☆31Updated last month
- CUDA implementation of autoregressive linear attention, with all the latest research findings☆44Updated 2 years ago
- Experiment of using Tangent to autodiff triton☆79Updated last year
- ☆53Updated 8 months ago
- RWKV model implementation☆38Updated last year
- Blog post☆17Updated last year
- Unofficially Implements https://arxiv.org/abs/2112.05682 to get Linear Memory Cost on Attention for PyTorch☆12Updated 3 years ago
- Serialize JAX, Flax, Haiku, or Objax model params with 🤗`safetensors`☆44Updated last year
- ☆29Updated 2 years ago
- ☆33Updated 8 months ago
- The official code of "Building on Efficient Foundations: Effectively Training LLMs with Structured Feedforward Layers"☆19Updated 10 months ago
- Code Release for "Broken Neural Scaling Laws" (BNSL) paper☆58Updated last year
- Official repository for the paper "Approximating Two-Layer Feedforward Networks for Efficient Transformers"☆37Updated last year
- Transformer with Mu-Parameterization, implemented in Jax/Flax. Supports FSDP on TPU pods.☆30Updated last week
- ☆12Updated 6 months ago
- Demo of the unit_scaling library, showing how a model can be easily adapted to train in FP8.☆44Updated 10 months ago