# Copyright 2026 The OpenXLA Authors. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. # ============================================================================== toc: - heading: XLA documentation - title: Getting started # This is the default tab for the Getting started section. section: - title: Introduction path: /xla - title: XLA architecture overview path: /xla/architecture - title: XLA terminology path: /xla/terminology - title: XLA flags guidance path: /xla/flags_guidance - title: Topical guides # This is the default tab for the Topical guides section. section: - title: Journey from HLO to executable path: /xla/hlo_to_thunks # XLA Ops sub-section - heading: XLA Ops - title: Operation semantics path: /xla/operation_semantics - title: HLO Passes path: /xla/hlo_passes - title: Aliasing path: /xla/aliasing - title: Async HLO instructions path: /xla/async_ops # Data representation sub-section - heading: Data representation - title: Shapes and layout path: /xla/shapes - title: Tiled layout path: /xla/tiled_layout - title: Symbolic expressions and maps path: /xla/symbolic_expression - title: Indexing Analysis path: /xla/indexing - title: Broadcasting path: /xla/broadcasting # GPU backend sub-section - heading: GPU backend - title: GPU architecture path: /xla/gpu_architecture - title: GPU Emitters path: /xla/emitters - title: Persisted autotuning path: /xla/persisted_autotuning - title: Determinism (GPU) path: /xla/determinism # TPU backend sub-section - heading: TPU backend - title: Sparsecore (embeddings) path: /xla/sparsecore - title: Megascale path: /xla/megascale/overview # Performance sub-section - heading: Performance - title: LHS Cost Model path: /xla/lhs_cost_model - title: Effort Levels path: /xla/effort_levels # Advanced guides sub-section - heading: Advanced guides - title: Write custom calls path: /xla/custom_call - title: Develop a new backend for XLA path: /xla/developing_new_backend - title: Debugging section: # This is the default tab for the Debugging section. - heading: XLA tooling - title: Dump HLO computations path: /xla/hlo_dumps - title: Errors overview path: /xla/errors_overview - title: Error codes path: /xla/error_codes - title: Debug OOM errors with XProf path: /xla/oom_debugging - title: Debug Megascale path: /xla/megascale/debugging_workflow # XLA Tools sub-section - heading: XLA tools - title: XLA tooling path: /xla/tools - title: Multi-host HLO runner path: /xla/tools_multihost_hlo_runner - title: Contributing # This is the default tab for the Contributing section. section: - title: Developer guide path: /xla/developer_guide - title: Set up development environment path: /xla/developer_environment - title: Build from source path: /xla/build_from_source - title: Testing HLO passes path: /xla/test_hlo_passes - title: Copybara quirks path: /xla/copybara - title: PJRT plugins # This is the default tab for the PJRT section. section: - title: "PJRT - Uniform Device API" path: /xla/pjrt/overview - title: PJRT C++ API overview path: /xla/pjrt/cpp_api_overview - title: Develop a new PJRT plugin path: /xla/pjrt/pjrt_integration - title: PJRT examples path: /xla/pjrt/examples - title: XLA frontends # This is the default tab for the XLA frontends section. section: - title: JAX path: https://docs.jax.dev/en/latest/notebooks/thinking_in_jax.html status: external - title: "PyTorch/XLA" path: https://docs.pytorch.org/xla/release/r2.8/index.html#pytorch-xla-documentation status: external - heading: TensorFlow - title: Use XLA in TensorFlow path: /xla/tf2xla - title: Autoclustering tutorial path: /xla/tf2xla/tutorials/autoclustering_xla - title: Use XLA with tf.function path: /xla/tf2xla/tutorials/jit_compile