.buckconfig
.buckroot
.gitignore
.pre-commit-config.yaml
BUCK
CODEOWNERS
LICENSE.txt
README.md
pyproject.toml
pytest.ini
requirements-dev.txt
requirements.txt
train.py
.agents/skills/verify-run-output/SKILL.md
.agents/skills/verify-run-output/reference.md
.agents/skills/verify-run-output/scripts/verify_run.py
.github/ci-constraints.txt
.github/workflows/cleanup_releases.yml
.github/workflows/cpu-tests.yml
.github/workflows/gemm-sweep-analysis.yml
.github/workflows/nightly.yml
.github/workflows/pre-commit.yml
.github/workflows/rccl-warp-speed-analysis.yml
.github/workflows/release.yml
analysis/overlap_report.py
analysis/figures/GEMM_Comm_CU_Contention.png
analysis/figures/baseline_timeline.png
analysis/figures/overlap_breakdown.png
analysis/figures/overlap_ratio.png
analysis/figures/sdma_benchmark.png
analysis/figures/training_bad_overlap.png
config/default.yaml
config/distributed.yaml
config/distributed_two_nodes.yaml
config/mi350_overlap_stress.yaml
config/profile_overlap_2gpu.yaml
config/profile_overlap_short.yaml
config/reproduce_hang.yaml
config/reproduce_hang_no_torch_prof.yaml
config/shampoo_opt.yaml
config/tf32_overlap_eval.yaml
config/multi_node/distributed_multinode.yaml
config/multi_node/shampoo_opt_multi_node.yaml
config/race/env_vars_reference.yaml
config/race/minimal_reproducer.yaml
config/single_node/gemm_overlap_comm.yaml
config/torchrec_dist/sparse_data_dist_base.yaml
config/torchrec_dist/sparse_data_dist_overlap_ineff.yaml
config/torchrec_dist/sparse_data_dist_overlap_latency.yaml
config/torchrec_dist/sparse_data_dist_queue_contended.yaml
docker/.env.example
docker/Dockerfile.rocm-ubuntu-ebpf
docker/Dockerfile.rocm70_2-ubuntu-nan
docker/Dockerfile.rocm70_2-ubuntu-pytorch
docker/Dockerfile.rocm70_9-1
docker/Dockerfile.rocm70_9-1-shampoo
docker/README.md
docker/docker-compose.build.yaml
docker/docker-compose.rccl.yaml
docker/docker-compose.rocm70_9-1-shampoo.yaml
docker/docker-compose.rocm70_9-1.yaml
docker/docker-compose.yaml
docker/setup-env.sh
docker/rccl_test/Dockerfile.rocm70_9-1
docker/rccl_test/docker-compose.rocm70_9-1.yaml
docs/buck2.md
docs/ci-testing-plan.md
docs/comprehensive_report.html
docs/configuration.md
docs/ebpf-integration.md
docs/ebpf-usage-guide.md
docs/env-probe.md
docs/getting-started.md
docs/hw-queue-eval.md
docs/hw_queue_cmds.png
docs/llm-determinism.md
docs/magpie-integration.md
docs/param_sweep.png
docs/profiling.md
docs/releasing.md
docs/running-benchmark.md
docs/troubleshooting.md
docs/agent/agentic-testing-guide.md
docs/agent/aorta-probe-agent.md
docs/plans/aorta-probe-188-rubric.md
docs/probe/bundle.md
docs/probe/classifier.md
docs/probe/handout-templates.md
docs/probe/redaction.md
docs/probe/sandbox.md
docs/probe/usage.md
examples/mitigations-sidecar.json
examples/probe-flag-sidecar.json
misc/smci350-zts-gtu-c6-25_200691.1758927066581847260.pt.trace.json
notebooks/aorta_training.ipynb
recipes/README-running-recipes.md
recipes/README.md
recipes/ainic-gdr-flush-sdc-quick.yaml
recipes/ainic-gdr-flush-sdc.yaml
recipes/ainic-sdc-stress-reproducer.yaml
recipes/ainic-smoke.yaml
recipes/example-fsdp-smoke.yaml
recipes/example-inference-continuous-smoke.yaml
recipes/example-inference-smoke.yaml
recipes/example-llm-determinism.yaml
recipes/example-probe-smoke.yaml
recipes/example-training-ddp-smoke.yaml
recipes/example-training-fsdp-smoke.yaml
recipes/probe-flag-sweep.yaml
recipes/probe-template-bash.yaml
recipes/probe-template-buck2.yaml
recipes/probe-template-torchrun.yaml
recipes/race_smoke.yaml
scripts/analyze_metrics.py
scripts/analyze_sweep_results.py
scripts/bump_version.py
scripts/check_wheel_version.py
scripts/debug_hang.sh
scripts/ebpf-entrypoint.sh
scripts/launch_cuda.sh
scripts/launch_rocm.sh
scripts/optuna_sweep.py
scripts/reproduce_hang.sh
scripts/reproduce_hang_with_rocprof.sh
scripts/rocprof_capture.sh
scripts/run_fsdp_sweep.sh
scripts/run_full_sweep.sh
scripts/run_quick_sweep.sh
scripts/run_sdma_prototype.py
scripts/run_with_hang_detection.sh
scripts/serve_report.py
scripts/stress_test_hang.sh
scripts/test_gpu_queue_impact.sh
scripts/tracelens_with_gemm_patch.py
scripts/visualize_timeline.py
scripts/gemm_analysis/README.md
scripts/gemm_analysis/analyze_gemm_reports.py
scripts/gemm_analysis/create_embeded_html_report.py
scripts/gemm_analysis/enhance_gemm_variance_with_timestamps.py
scripts/gemm_analysis/gemm_report_with_collective_overlap.py
scripts/gemm_analysis/html_template.py
scripts/gemm_analysis/plot_gemm_variance.py
scripts/gemm_analysis/process_comms.py
scripts/gemm_analysis/process_gpu_timeline.py
scripts/gemm_analysis/rocprof_cu_only.yaml
scripts/gemm_analysis/rocprof_input.yaml
scripts/gemm_analysis/run_tracelens_analysis.sh
scripts/gemm_analysis/run_train_various_channels.sh
scripts/hw_queue/compare_baselines.py
scripts/hw_queue/profile_queues.sh
scripts/hw_queue/run_sweep.sh
scripts/multi_node/README.md
scripts/multi_node/config_node.sh
scripts/multi_node/launch_reproducer.sh
scripts/multi_node/local_launch.sh
scripts/multi_node/master_launch.sh
scripts/multi_node/set_env_variables.sh
scripts/multi_node/setup_multi_node.sh
scripts/multi_node/start_docker_all_nodes.sh
scripts/rocm_mem_debug/README.md
scripts/rocm_mem_debug/rocm_mem_debug.py
scripts/rocm_mem_debug/tests/Makefile
scripts/rocm_mem_debug/tests/run_tests.sh
scripts/rocm_mem_debug/tests/negative/01_normal_alloc.cpp
scripts/rocm_mem_debug/tests/negative/02_synced_streams.cpp
scripts/rocm_mem_debug/tests/negative/03_correct_h2d_d2h.cpp
scripts/rocm_mem_debug/tests/negative/04_large_clean_alloc.cpp
scripts/rocm_mem_debug/tests/positive/01_use_after_free.cpp
scripts/rocm_mem_debug/tests/positive/02_oob_write.cpp
scripts/rocm_mem_debug/tests/positive/03_null_ptr_deref.cpp
scripts/rocm_mem_debug/tests/positive/04_eviction_storm.cpp
scripts/rocm_mem_debug/tests/positive/05_unsynced_streams.cpp
scripts/rocm_mem_debug/tests/positive/06_large_alloc_copy.cpp
scripts/tracelens_single_config/README.md
scripts/tracelens_single_config/add_collective_comparison.py
scripts/tracelens_single_config/add_comparison_sheets.py
scripts/tracelens_single_config/combine_reports.py
scripts/tracelens_single_config/compare_all_runs.py
scripts/tracelens_single_config/create_final_html.py
scripts/tracelens_single_config/create_final_plots.py
scripts/tracelens_single_config/create_final_report.py
scripts/tracelens_single_config/html_report_config.py
scripts/tracelens_single_config/process_gpu_timeline.py
scripts/tracelens_single_config/run_full_analysis.py
scripts/tracelens_single_config/run_rccl_warp_speed_comparison.sh
scripts/tracelens_single_config/run_tracelens_single_config.sh
scripts/utils/html_to_md.py
scripts/utils/merge_gpu_trace_ranks.py
src/amd_aorta.egg-info/PKG-INFO
src/amd_aorta.egg-info/SOURCES.txt
src/amd_aorta.egg-info/dependency_links.txt
src/amd_aorta.egg-info/entry_points.txt
src/amd_aorta.egg-info/requires.txt
src/amd_aorta.egg-info/top_level.txt
src/aorta/__init__.py
src/aorta/agent/__init__.py
src/aorta/agent/llm.py
src/aorta/agent/loop.py
src/aorta/agent/policy.py
src/aorta/agent/report.py
src/aorta/agent/state.py
src/aorta/bundle/__init__.py
src/aorta/bundle/_slug.py
src/aorta/bundle/errors.py
src/aorta/bundle/manifest.py
src/aorta/bundle/redactor.py
src/aorta/bundle/writer.py
src/aorta/cli/__init__.py
src/aorta/cli/_deprecation.py
src/aorta/cli/agent.py
src/aorta/cli/bench.py
src/aorta/cli/bundle.py
src/aorta/cli/env.py
src/aorta/cli/environments.py
src/aorta/cli/mitigations.py
src/aorta/cli/probe.py
src/aorta/cli/run.py
src/aorta/cli/sweep.py
src/aorta/cli/triage.py
src/aorta/data/__init__.py
src/aorta/data/synthetic_dataset.py
src/aorta/ebpf/__init__.py
src/aorta/ebpf/events.py
src/aorta/ebpf/parser.py
src/aorta/ebpf/runner.py
src/aorta/ebpf/scripts/PROVENANCE.md
src/aorta/ebpf/scripts/gpu_cont.bt
src/aorta/ebpf/scripts/gpu_cont_1kprobe.bt
src/aorta/ebpf/scripts/gpu_cont_light.bt
src/aorta/ebpf/scripts/gpu_cont_tp_only.bt
src/aorta/ebpf/scripts/gpu_cont_unrelated_kprobe.bt
src/aorta/experiments/__init__.py
src/aorta/experiments/sdma_prototype.py
src/aorta/hw_queue_eval/__init__.py
src/aorta/hw_queue_eval/__main__.py
src/aorta/hw_queue_eval/cli.py
src/aorta/hw_queue_eval/core/__init__.py
src/aorta/hw_queue_eval/core/device_ebpf.py
src/aorta/hw_queue_eval/core/ebpf_dma_tracer.py
src/aorta/hw_queue_eval/core/ebpf_memory_tracer.py
src/aorta/hw_queue_eval/core/ebpf_nan_correlator.py
src/aorta/hw_queue_eval/core/ebpf_race_detector.py
src/aorta/hw_queue_eval/core/ebpf_rccl_tracer.py
src/aorta/hw_queue_eval/core/ebpf_tracer.py
src/aorta/hw_queue_eval/core/harness.py
src/aorta/hw_queue_eval/core/metrics.py
src/aorta/hw_queue_eval/core/policy_evaluator.py
src/aorta/hw_queue_eval/core/profiler.py
src/aorta/hw_queue_eval/core/torch_profiler.py
src/aorta/hw_queue_eval/workloads/__init__.py
src/aorta/hw_queue_eval/workloads/base.py
src/aorta/hw_queue_eval/workloads/registry.py
src/aorta/hw_queue_eval/workloads/distributed/__init__.py
src/aorta/hw_queue_eval/workloads/distributed/activation_ckpt.py
src/aorta/hw_queue_eval/workloads/distributed/comms_compute_overlap.py
src/aorta/hw_queue_eval/workloads/distributed/fsdp_tp.py
src/aorta/hw_queue_eval/workloads/distributed/grad_accum.py
src/aorta/hw_queue_eval/workloads/distributed/moe.py
src/aorta/hw_queue_eval/workloads/inference/__init__.py
src/aorta/hw_queue_eval/workloads/inference/continuous_batch.py
src/aorta/hw_queue_eval/workloads/inference/rag_pipeline.py
src/aorta/hw_queue_eval/workloads/inference/speculative_decode.py
src/aorta/hw_queue_eval/workloads/latency_sensitive/__init__.py
src/aorta/hw_queue_eval/workloads/latency_sensitive/graph_subgraphs.py
src/aorta/hw_queue_eval/workloads/latency_sensitive/hetero_kernels.py
src/aorta/hw_queue_eval/workloads/pipeline/__init__.py
src/aorta/hw_queue_eval/workloads/pipeline/async_dataload.py
src/aorta/hw_queue_eval/workloads/pipeline/torch_compile.py
src/aorta/hw_queue_eval/workloads/pipeline/zero_offload.py
src/aorta/instrumentation/README.md
src/aorta/instrumentation/__init__.py
src/aorta/instrumentation/_probe_main.py
src/aorta/instrumentation/buck_introspect.py
src/aorta/instrumentation/build_system.py
src/aorta/instrumentation/checksum.py
src/aorta/instrumentation/determinism.py
src/aorta/instrumentation/environment.py
src/aorta/instrumentation/layer_numerics/README.md
src/aorta/instrumentation/layer_numerics/__init__.py
src/aorta/instrumentation/layer_numerics/__main__.py
src/aorta/instrumentation/layer_numerics/instrument_nan_logger.py
src/aorta/instrumentation/recipes/__init__.py
src/aorta/instrumentation/recipes/buck.py
src/aorta/models/__init__.py
src/aorta/models/ranking_transformer.py
src/aorta/models/repeated_block.py
src/aorta/probe/__init__.py
src/aorta/probe/bundle_hook.py
src/aorta/probe/cli_helpers.py
src/aorta/probe/recipe_builder.py
src/aorta/probe/redaction.py
src/aorta/probe/resume.py
src/aorta/probe/sandbox.py
src/aorta/probe/classifier/__init__.py
src/aorta/probe/classifier/disables.py
src/aorta/probe/classifier/tier1_process.py
src/aorta/probe/classifier/tier2_hang.py
src/aorta/probe/classifier/tier3_kernel.py
src/aorta/probe/classifier/tier4_patterns.py
src/aorta/probe/classifier/tier5_custom.py
src/aorta/probe/classifier/verdict.py
src/aorta/profiling/__init__.py
src/aorta/profiling/kernel_profiler.py
src/aorta/profiling/stream_profiler.py
src/aorta/race/README.md
src/aorta/race/__init__.py
src/aorta/race/__main__.py
src/aorta/race/background.md
src/aorta/race/base.py
src/aorta/race/compute.py
src/aorta/race/config.py
src/aorta/race/developer_guide.md
src/aorta/race/modes/__init__.py
src/aorta/race/modes/ddp.py
src/aorta/race/modes/default.py
src/aorta/race/modes/fsdp.py
src/aorta/registry/README.md
src/aorta/registry/__init__.py
src/aorta/registry/environments.py
src/aorta/registry/errors.py
src/aorta/registry/mitigations.py
src/aorta/registry/sidecar.py
src/aorta/registry/types.py
src/aorta/report/README.md
src/aorta/report/__init__.py
src/aorta/report/__main__.py
src/aorta/report/aorta-report-ug.md
src/aorta/report/cli.py
src/aorta/report/magpie_adapter.py
src/aorta/report/utils.py
src/aorta/report/analysis/__init__.py
src/aorta/report/analysis/analyze_gemm.py
src/aorta/report/analysis/analyze_single.py
src/aorta/report/analysis/analyze_sweep.py
src/aorta/report/analysis/cli.py
src/aorta/report/analysis/kernel_correlator.py
src/aorta/report/analysis/tracelens_wrapper.py
src/aorta/report/comparison/__init__.py
src/aorta/report/comparison/cli.py
src/aorta/report/comparison/collective_comparison.py
src/aorta/report/comparison/combine.py
src/aorta/report/comparison/formatting.py
src/aorta/report/comparison/gpu_timeline_comparison.py
src/aorta/report/generators/__init__.py
src/aorta/report/generators/cli.py
src/aorta/report/generators/excel_report.py
src/aorta/report/generators/gemm_report.py
src/aorta/report/generators/html_generator.py
src/aorta/report/generators/kernel_report.py
src/aorta/report/generators/performance_report.py
src/aorta/report/generators/plot_generator.py
src/aorta/report/generators/sweep_comparison.py
src/aorta/report/generators/plot_helper/__init__.py
src/aorta/report/generators/plot_helper/common.py
src/aorta/report/generators/plot_helper/gemm_boxplots.py
src/aorta/report/generators/plot_helper/gemm_data.py
src/aorta/report/generators/plot_helper/gemm_interaction.py
src/aorta/report/generators/plot_helper/gemm_violin.py
src/aorta/report/generators/plot_helper/gpu_by_rank.py
src/aorta/report/generators/plot_helper/gpu_heatmap.py
src/aorta/report/generators/plot_helper/gpu_percent_change.py
src/aorta/report/generators/plot_helper/nccl_charts.py
src/aorta/report/generators/plot_helper/summary_dashboard.py
src/aorta/report/pipelines/__init__.py
src/aorta/report/pipelines/cli.py
src/aorta/report/pipelines/gemm_pipeline.py
src/aorta/report/pipelines/summary_pipeline.py
src/aorta/report/processing/__init__.py
src/aorta/report/processing/cli.py
src/aorta/report/processing/gpu_timeline_single.py
src/aorta/report/processing/gpu_timeline_sweep.py
src/aorta/report/processing/process_comms.py
src/aorta/report/processing/process_gemm_variance.py
src/aorta/report/templates/__init__.py
src/aorta/report/templates/gemm_report_template.py
src/aorta/report/templates/performance_report_template.py
src/aorta/report/templates/sweep_comparison_template.py
src/aorta/run/__init__.py
src/aorta/run/cli_helpers.py
src/aorta/run/collectors.py
src/aorta/run/discovery.py
src/aorta/run/dispatcher.py
src/aorta/run/results.py
src/aorta/run/retention.py
src/aorta/run/validation.py
src/aorta/training/__init__.py
src/aorta/training/fsdp_trainer.py
src/aorta/triage/__init__.py
src/aorta/triage/confound.py
src/aorta/triage/matrix.py
src/aorta/triage/output.py
src/aorta/triage/recipe.py
src/aorta/triage/runner.py
src/aorta/utils/__init__.py
src/aorta/utils/config.py
src/aorta/utils/device.py
src/aorta/utils/distributed.py
src/aorta/utils/gpu_control.py
src/aorta/utils/logging.py
src/aorta/utils/streams.py
src/aorta/utils/timing.py
src/aorta/workloads/__init__.py
src/aorta/workloads/_base.py
src/aorta/workloads/_subprocess.py
src/aorta/workloads/inference.py
src/aorta/workloads/llm_determinism.py
src/aorta/workloads/race.py
src/aorta/workloads/training.py
tests/__init__.py
tests/conftest.py
tests/test_bump_version.py
tests/test_check_wheel_version.py
tests/test_distribution_name.py
tests/test_gpu_control.py
tests/test_merge_gpu_trace_ranks.py
tests/agent/test_cli.py
tests/agent/test_llm.py
tests/agent/test_loop.py
tests/agent/test_outcomes.py
tests/agent/test_policy.py
tests/agent/test_report.py
tests/agent/test_state.py
tests/bundle/__init__.py
tests/bundle/conftest.py
tests/bundle/test_cli.py
tests/bundle/test_manifest.py
tests/bundle/test_no_new_deps.py
tests/bundle/test_writer.py
tests/cli/__init__.py
tests/cli/test_bench.py
tests/ebpf/__init__.py
tests/ebpf/test_events.py
tests/ebpf/test_parser.py
tests/ebpf/test_runner.py
tests/fixtures/buck/audit_dependencies.json
tests/hw_queue_eval/__init__.py
tests/hw_queue_eval/test_ebpf_nan_debugging.py
tests/hw_queue_eval/test_ebpf_tracer.py
tests/hw_queue_eval/test_harness.py
tests/hw_queue_eval/test_metrics.py
tests/hw_queue_eval/test_workloads.py
tests/instrumentation/__init__.py
tests/instrumentation/test_buck_introspect.py
tests/instrumentation/test_buck_recipe.py
tests/instrumentation/test_build_system.py
tests/instrumentation/test_checksum.py
tests/instrumentation/test_environment.py
tests/instrumentation/test_layer_numerics.py
tests/instrumentation/test_recipe_cli.py
tests/probe/__init__.py
tests/probe/test_bundle_integration.py
tests/probe/test_cli_parsing.py
tests/probe/test_dry_run.py
tests/probe/test_end_to_end.py
tests/probe/test_env_passthrough.py
tests/probe/test_handout_templates.py
tests/probe/test_list_patterns.py
tests/probe/test_probe_env_noise_e2e.py
tests/probe/test_recipe.py
tests/probe/test_recurring_issue_regression.py
tests/probe/test_redaction.py
tests/probe/test_resume.py
tests/probe/test_retention.py
tests/probe/test_sandbox.py
tests/probe/test_shared_engine.py
tests/probe/test_subprocess_workload.py
tests/probe/classifier/__init__.py
tests/probe/classifier/test_disables.py
tests/probe/classifier/test_tier1.py
tests/probe/classifier/test_tier2.py
tests/probe/classifier/test_tier3.py
tests/probe/classifier/test_tier4.py
tests/probe/classifier/test_tier5.py
tests/probe/classifier/test_verdict.py
tests/probe/fixtures/probe_minimal.yaml
tests/probe/fixtures/probe_needs_sidecar.yaml
tests/probe/fixtures/probe_needs_two_sidecars.yaml
tests/probe/fixtures/probe_sidecar.json
tests/probe/fixtures/probe_sidecar_two.json
tests/probe/fixtures/probe_two_cell.yaml
tests/probe/fixtures/probe_with_phase_2_keys.yaml
tests/probe/fixtures/redaction_input.txt
tests/probe/fixtures/conditions/hostile.txt
tests/probe/fixtures/tier4_logs/collective_timeout.txt
tests/probe/fixtures/tier4_logs/cuda_error.txt
tests/probe/fixtures/tier4_logs/hip_error.txt
tests/probe/fixtures/tier4_logs/nan_signature.txt
tests/probe/fixtures/tier4_logs/nccl_rccl_error.txt
tests/probe/fixtures/tier4_logs/python_traceback.txt
tests/probe/fixtures/tier4_logs/rocm_error.txt
tests/profiling/__init__.py
tests/profiling/test_fsdp_trainer_kernel_profiler_lifecycle.py
tests/profiling/test_kernel_profiler.py
tests/registry/__init__.py
tests/registry/conftest.py
tests/registry/test_environments.py
tests/registry/test_mitigations.py
tests/registry/test_sidecar.py
tests/report/__init__.py
tests/report/test_kernel_correlator.py
tests/report/test_kernel_report.py
tests/report/test_kernel_report_dependency_free.py
tests/run/__init__.py
tests/run/test_cli_helpers.py
tests/run/test_cli_parsing.py
tests/run/test_discovery.py
tests/run/test_dispatcher.py
tests/run/test_integration.py
tests/run/test_results.py
tests/run/test_retention.py
tests/run/test_validation.py
tests/sweep/__init__.py
tests/sweep/test_sweep_cli.py
tests/triage/__init__.py
tests/triage/test_confound.py
tests/triage/test_matrix.py
tests/triage/test_matrix_error_column.py
tests/triage/test_matrix_top_columns.py
tests/triage/test_output_layout.py
tests/triage/test_recipe.py
tests/triage/test_run_dir_lock.py
tests/triage/test_run_summary.py
tests/triage/test_runner_b1_api.py
tests/triage/test_stop_after.py
tests/workloads/__init__.py
tests/workloads/test_inference.py
tests/workloads/test_llm_determinism.py
tests/workloads/test_race.py
tests/workloads/test_race_checksums.py
tests/workloads/test_race_real_backward.py
tests/workloads/test_race_transformer_smoke.py
tests/workloads/test_training.py
third-party/python/BUCK
toolchains/BUCK