mirrors/rocm_jax
1
0
Fork 0
mirror of https://github.com/ROCm/jax.git synced 2025-04-17 04:16:07 +00:00
Code Issues Packages Projects Releases Wiki Activity
rocm_jax/jaxlib
History
Tomás Longeri 20d9aac6be [Mosaic] Remove some restrictions for vector.shape_cast in infer-vector-layout and apply-vector-layout 
- On infer-vector-layout remove some restrictions related to batch dimensions. Reshaping them doesn't matter as long as they don't combine with tiled dimensions.
- On apply-vector-layout, simplify handling of cases where the implicit tiled don't change while removing some unnecessary restrictions.
  - Don't require native tiling or natural topology for this.

PiperOrigin-RevId: 640837740
2024-06-06 03:26:43 -07:00
..
cpu
Remove DUCC FFT from jaxlib
2024-05-28 21:12:23 -07:00
cuda
[Mosaic GPU] Integrate the ExecutionEngine with the jaxlib GPU plugin
2024-05-30 01:46:23 -07:00
gpu
add cholesky changes in bazel
2024-05-18 00:37:09 +00:00
mlir
[Mosaic GPU] Integrate the ExecutionEngine with the jaxlib GPU plugin
2024-05-30 01:46:23 -07:00
mosaic
[Mosaic] Remove some restrictions for vector.shape_cast in infer-vector-layout and apply-vector-layout
2024-06-06 03:26:43 -07:00
rocm
add cholesky changes in bazel
2024-05-18 00:37:09 +00:00
tools
[Mosaic GPU] Integrate the ExecutionEngine with the jaxlib GPU plugin
2024-05-30 01:46:23 -07:00
triton
Removed inspect.signature() call from jaxlib.triton.dialect.ScanOp
2024-03-11 13:30:41 -07:00
absl_status_casters.h
Switch jaxlib to use nanobind instead of pybind11.
2023-08-24 16:07:56 -07:00
BUILD
Expose XLA FFI headers to bazel build and re-enable tests
2024-06-04 10:14:43 -07:00
cpu_feature_guard.c
Add a comment about x86 builds of Python to cpu_feature_guard
2024-05-29 15:25:44 -04:00
cuda_plugin_extension.cc
[xla:python] Add support for registering custom call targets for all XLA execution stages and for XLA FFI traits
2024-05-24 10:34:46 -07:00
ducc_fft.py
Add python stub files for jaxlib/cpu C++ Python extensions.
2023-11-28 08:45:24 -08:00
gpu_common_utils.py
Improve the gpu lowering error message if users forget link the gpu library.
2023-09-11 16:14:18 -07:00
gpu_linalg.py
Cholesky rank-1 update kernel for JAX.
2024-05-14 15:21:38 -07:00
gpu_prng.py
Remove support for CUDA 11.
2024-03-25 11:46:39 -07:00
gpu_rnn.py
Remove support for CUDA 11.
2024-03-25 11:46:39 -07:00
gpu_solver.py
Clean up version switches from dense array migration
2024-05-28 10:58:51 -07:00
gpu_sparse.py
Remove support for CUDA 11.
2024-03-25 11:46:39 -07:00
gpu_triton.py
Remove support for CUDA 11.
2024-03-25 11:46:39 -07:00
handle_pool.h
[GPU] Add missing stream synchronization to tridiagonal_solve gtsv2 call.
2023-05-08 15:37:04 -07:00
hlo_helpers.py
Clean up version switches from dense array migration
2024-05-28 10:58:51 -07:00
init.py
Change JAX's copyright attribution from "Google LLC" to "The JAX Authors.".
2022-09-22 12:27:19 -07:00
jax.bzl
Add data argument to jax_test Bazel rule, forward to py_test
2024-06-04 11:17:30 -04:00
kernel_helpers.h
Change JAX's copyright attribution from "Google LLC" to "The JAX Authors.".
2022-09-22 12:27:19 -07:00
kernel_nanobind_helpers.h
Move tsl/python to xla/tsl/python
2024-03-29 13:15:21 -07:00
lapack.py
Use MLIR generated convenience functions athing(...) instead of writing AThingOp(...).result.
2023-11-17 11:47:14 -08:00
README.md
Populate long_description for jax & jaxlib
2022-07-13 14:03:32 -07:00
setup.py
Update cuDNN to version 9.1.1 in JAX
2024-05-24 10:10:21 -07:00
symlink_files.bzl
Change JAX's copyright attribution from "Google LLC" to "The JAX Authors.".
2022-09-22 12:27:19 -07:00
utils.cc
Switch jaxlib to use nanobind instead of pybind11.
2023-08-24 16:07:56 -07:00

README.md

jaxlib: support library for JAX

jaxlib is the support library for JAX. While JAX itself is a pure Python package, jaxlib contains the binary (C/C++) parts of the library, including Python bindings, the XLA compiler, the PJRT runtime, and a handful of handwritten kernels. For more information, including installation and build instructions, refer to main JAX README: https://github.com/google/jax/.