Two implementations of ZeRO-1 optimizer sharding in JAX
-
Updated
Jun 11, 2023 - Python
#
pjit
Here are 3 public repositories matching this topic...
Tensor Parallelism with JAX + Shard Map
-
Updated
Sep 29, 2023 - Python
Comprehensive JAX implementation of neural networks and scientific computing. Features distributed training, physics-informed networks, custom autodiff, and advanced optimization. Production-ready code with numerical stability, multi-device parallelism, and research-grade implementations.
machine-learning research deep-learning optimization automatic-differentiation parallel-computing transformers bayesian-methods scientific-computing neural-networks high-performance-computing differential-equations gpu-computing pmap distributed-training jax mixed-precision physics-informed pjit tpu-computing
-
Updated
Mar 9, 2026 - Jupyter Notebook
Add this topic to your repo
To associate your repository with the pjit topic, visit your repo's landing page and select "manage topics."