arogozhnikov / adamw_bfloat16Links
AdamW optimizer for bfloat16 models in pytorch 🔥.
☆38Updated last year
Alternatives and similar repositories for adamw_bfloat16
Users that are interested in adamw_bfloat16 are comparing it to the libraries listed below
Sorting:
- ☆32Updated last year
- Latent Diffusion Language Models☆70Updated 2 years ago
- Implementation of Token Shift GPT - An autoregressive model that solely relies on shifting the sequence space for mixing☆50Updated 3 years ago
- Engineering the state of RNN language models (Mamba, RWKV, etc.)☆32Updated last year
- A place to store reusable transformer components of my own creation or found on the interwebs☆63Updated last week
- RWKV model implementation☆38Updated 2 years ago
- Standalone Product Key Memory module in Pytorch - for augmenting Transformer models☆87Updated last month
- Triton Implementation of HyperAttention Algorithm☆48Updated 2 years ago
- CUDA implementation of autoregressive linear attention, with all the latest research findings☆46Updated 2 years ago
- ☆31Updated last month
- AdaCat☆49Updated 3 years ago
- Implementation of a Transformer using ReLA (Rectified Linear Attention) from https://arxiv.org/abs/2104.07012☆49Updated 3 years ago
- ☆29Updated 3 years ago
- Explorations into adversarial losses on top of autoregressive loss for language modeling☆38Updated last week
- Using FlexAttention to compute attention with different masking patterns☆47Updated last year
- LayerNorm(SmallInit(Embedding)) in a Transformer to improve convergence☆61Updated 3 years ago
- MaskedTensors for PyTorch☆38Updated 3 years ago
- Unofficially Implements https://arxiv.org/abs/2112.05682 to get Linear Memory Cost on Attention for PyTorch☆12Updated 3 years ago
- My explorations into editing the knowledge and memories of an attention network☆35Updated 3 years ago
- Official Repository of Pretraining Without Attention (BiGS), BiGS is the first model to achieve BERT-level transfer learning on the GLUE …☆116Updated last year
- Demo of the unit_scaling library, showing how a model can be easily adapted to train in FP8.☆46Updated last year
- Implementation of the Kalman Filtering Attention proposed in "Kalman Filtering Attention for User Behavior Modeling in CTR Prediction"☆59Updated 2 years ago
- A python library for highly configurable transformers - easing model architecture search and experimentation.☆49Updated 4 years 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 2 years ago
- A case study of efficient training of large language models using commodity hardware.☆68Updated 3 years ago
- High performance pytorch modules☆18Updated 2 years ago
- Blog post☆17Updated last year
- Fast, Modern, and Low Precision PyTorch Optimizers☆118Updated 3 months ago
- ☆82Updated last year
- Automatically take good care of your preemptible TPUs☆37Updated 2 years ago