.gitignore
ANE_POLICY.md
CHANGELOG.md
CITATION.cff
CONTRIBUTING.md
LICENSE
MANIFEST.in
NOTICE
ORIGIN.json
QUICKSTART.md
README.md
README.zh-CN.md
RELEASE_STATUS.json
SECURITY.md
build_release.py
package-manifest.json
package_release.py
pyproject.toml
requirements-runtime.txt
setup.py
validate_package.py
config/nodes/m4.json
config/nodes/m5.json
config/nodes/m6.json
docs/ANE_DOWNLOADS.md
docs/ANE_PROVISIONING.md
docs/BRANDING.md
docs/CONFIGURATION.md
docs/DEPLOYMENT.md
docs/IMPLEMENTATION.md
docs/PREVIEW_VALIDATION.json
docs/PYPI.md
docs/RESULTS.md
docs/RUNTIME_DISTRIBUTION.md
docs/THINKING_BUDGET.md
docs/VALIDATION.md
docs/system-comparison.csv
docs/assets/justfit-banner-mobile.svg
docs/assets/justfit-banner.svg
docs/assets/system-comparison.png
paper/JustFit-v2-arxiv-source.zip
paper/JustFit-v2.pdf
paper/README.md
paper/citation.bib
python/justfit_llm/__init__.py
python/justfit_llm/ane_bundle.py
python/justfit_llm/cli.py
python/justfit_llm/runtime_bundle.py
python/justfit_llm.egg-info/PKG-INFO
python/justfit_llm.egg-info/SOURCES.txt
python/justfit_llm.egg-info/dependency_links.txt
python/justfit_llm.egg-info/entry_points.txt
python/justfit_llm.egg-info/requires.txt
python/justfit_llm.egg-info/top_level.txt
scripts/check_ane_routes.py
scripts/check_broker_history.py
scripts/check_node_budget.py
scripts/check_thinking_budget.py
scripts/package_ane.py
scripts/package_runtime.py
scripts/qualify_m4_build.py
scripts/qualify_pip.py
scripts/run_local.py
service/http_guard.py
service/managed_base.py
service/requirements.txt
service/service.py
service/storage_guard.py
service/supervisor.py
service/test_disconnect.py
service/test_security.py
service/test_service.py
service/test_storage.py
service/test_supervisor_ports.py
service/broker-integration/BASE_SHA256.json
service/broker-integration/README.md
service/broker-integration/history-preservation.patch
service/broker-integration/pool-probe-lock.patch
service/broker-integration/test_pool_probe_lock.py
source/CONTRIBUTING.md
source/DEVELOPMENT.md
source/LICENSE
source/Makefile
source/README.md
source/SECURITY.md
source/THIRD_PARTY_NOTICES
source/justfit
source/dev/Makefile
source/dev/__init__.py
source/dev/evo4.env.example
source/dev/evo4.mk
source/dev/native.mk
source/dev/tests/engine/barrier_scheduler_test.cc
source/dev/tests/engine/cpu_affine_gpu_probe.cc
source/dev/tests/engine/cpu_affine_reference.cc
source/dev/tests/engine/cpu_host_probe.cc
source/dev/tests/engine/cpu_join_test.cc
source/dev/tests/engine/cpu_linear_reference.cc
source/dev/tests/engine/evo4_attention_probe.cc
source/dev/tests/engine/evo4_lookup_test.cc
source/dev/tests/engine/evo4_sampling_probe.cc
source/dev/tests/engine/explicit_budget_test.cc
source/dev/tests/engine/host_embedding_test.cc
source/dev/tests/engine/host_reserve_probe.cc
source/dev/tests/engine/prefill_alias_layout_test.cc
source/dev/tests/engine/qwen_state_storage_test.mm
source/dev/tests/engine/tq4_attention_probe.cc
source/dev/tests/engine/tq4_coalesced_bench.cc
source/dev/tests/engine/tq4_coalesced_warm_bench.cc
source/dev/tests/engine/tq4_compact_bench.cc
source/dev/tests/engine/tq4_dense_attention_probe.cc
source/dev/tests/engine/tq4_extent_layout_test.cc
source/dev/tests/engine/tq4_extent_storage_probe.cc
source/dev/tests/engine/tq4_staged_attention_probe.cc
source/dev/tests/engine/tq4_staged_multitile_probe.cc
source/dev/tests/engine/tq4_verify_boundary_probe.cc
source/dev/tests/engine/verify_a8_probe.cc
source/dev/tests/engine/weight_phase_probe.cc
source/dev/tools/__init__.py
source/dev/tools/build_config.py
source/dev/tools/build_identity.py
source/dev/tools/check_architecture.py
source/dev/tools/install.sh
source/dev/tools/installer_restarts.py
source/dev/tools/kernel_identity.py
source/dev/tools/package.py
source/dev/tools/package_brew.py
source/dev/tools/publish_test.py
source/dev/tools/update_model_catalog.py
source/dev/tools/vision_reference.py
source/dev/tools/weight_digests.mm
source/dev/tuning/AttentionFixture.hpp
source/dev/tuning/HostKvExtents.hpp
source/dev/tuning/LinearNumerics.hpp
source/dev/tuning/LinearTuning.cpp
source/dev/tuning/LinearTuning.hpp
source/dev/tuning/Measurement.cpp
source/dev/tuning/Measurement.hpp
source/dev/tuning/Tuning.cpp
source/dev/tuning/Tuning.hpp
source/dev/tuning/TuningWorkloads.cpp
source/dev/tuning/TuningWorkloads.hpp
source/dev/tuning/tune_kernels.mm
source/docs/performance.md
source/install/__init__.py
source/install/assembly.py
source/install/catalog.py
source/install/clients.py
source/install/families.py
source/install/gguf.py
source/install/hub.py
source/install/launcher.py
source/install/legacy.py
source/install/models.py
source/install/paths.py
source/install/requirements.txt
source/install/upstream.py
source/install/completions/_justfit
source/install/completions/justfit.bash
source/install/completions/justfit.fish
source/install/completions/models
source/install/completions/official-models.txt
source/install/completions/suggested-models.txt
source/runtime/AwakeClock.hpp
source/runtime/Checked.hpp
source/runtime/StderrLine.hpp
source/runtime/main.mm
source/runtime/ane/Handoff.hpp
source/runtime/ane/Handoff.mm
source/runtime/ane/Program.hpp
source/runtime/ane/Program.mm
source/runtime/ane/ProgramInstrumentation.hpp
source/runtime/engine/AneFfnStartup.cpp
source/runtime/engine/AneFfnStartup.hpp
source/runtime/engine/BenchmarkBarrier.hpp
source/runtime/engine/Bootstrap.hpp
source/runtime/engine/Bootstrap.mm
source/runtime/engine/Cache.cpp
source/runtime/engine/Cache.hpp
source/runtime/engine/CacheDirectory.cpp
source/runtime/engine/CacheDirectory.hpp
source/runtime/engine/CacheRecency.hpp
source/runtime/engine/DiskLabels.hpp
source/runtime/engine/Engine.cpp
source/runtime/engine/Engine.hpp
source/runtime/engine/FdTransport.cpp
source/runtime/engine/FdTransport.hpp
source/runtime/engine/Json.hpp
source/runtime/engine/KvCache.cpp
source/runtime/engine/KvCache.hpp
source/runtime/engine/KvPageTier.cpp
source/runtime/engine/KvPageTier.hpp
source/runtime/engine/KvPool.cpp
source/runtime/engine/KvPool.hpp
source/runtime/engine/KvTier.hpp
source/runtime/engine/MemoryAudit.cpp
source/runtime/engine/MemoryAudit.hpp
source/runtime/engine/MemoryControl.cpp
source/runtime/engine/MemoryControl.hpp
source/runtime/engine/MemoryGovernor.cpp
source/runtime/engine/MemoryGovernor.hpp
source/runtime/engine/MemoryPlan.cpp
source/runtime/engine/MemoryPlan.hpp
source/runtime/engine/NativeArguments.cpp
source/runtime/engine/NativeArguments.hpp
source/runtime/engine/NativeRuntime.cpp
source/runtime/engine/NativeRuntime.hpp
source/runtime/engine/Protocol.cpp
source/runtime/engine/Protocol.hpp
source/runtime/engine/RecencyOrder.hpp
source/runtime/engine/ReleasableMemory.hpp
source/runtime/engine/RuntimeResources.hpp
source/runtime/engine/RuntimeResources.mm
source/runtime/engine/Scheduler.cpp
source/runtime/engine/Scheduler.hpp
source/runtime/engine/StateCache.cpp
source/runtime/engine/StateCache.hpp
source/runtime/engine/Status.cpp
source/runtime/engine/Status.hpp
source/runtime/engine/Types.hpp
source/runtime/engine/WriteBehind.cpp
source/runtime/engine/WriteBehind.hpp
source/runtime/metal/BackendInstrumentation.hpp
source/runtime/metal/CommandGraph.hpp
source/runtime/metal/CommandWatchdog.hpp
source/runtime/metal/DeviceCapabilities.cpp
source/runtime/metal/DeviceCapabilities.hpp
source/runtime/metal/HostWork.hpp
source/runtime/metal/MetalBackend.hpp
source/runtime/metal/MetalBackend.mm
source/runtime/metal/Residency.hpp
source/runtime/metal/abi/AneFfn.h
source/runtime/metal/abi/DraftAttention.h
source/runtime/metal/abi/Embedding.h
source/runtime/metal/abi/ExecutionGeometry.h
source/runtime/metal/abi/GDN.h
source/runtime/metal/abi/Gguf.h
source/runtime/metal/abi/GgufRepack.h
source/runtime/metal/abi/KernelABI.h
source/runtime/metal/abi/KvExtent.h
source/runtime/metal/abi/Linear.h
source/runtime/metal/abi/MoE.h
source/runtime/metal/abi/PagedAttention.h
source/runtime/metal/abi/QuantFormat.h
source/runtime/metal/abi/QuantTables.h
source/runtime/metal/abi/RoPE.h
source/runtime/metal/abi/RowCopy.h
source/runtime/metal/abi/Sampling.h
source/runtime/metal/abi/Vision.h
source/runtime/metal/kernels/common/activation.h
source/runtime/metal/kernels/common/attention_gate.h
source/runtime/metal/kernels/common/attention_qkv_prepare.h
source/runtime/metal/kernels/common/draft_context_kv.h
source/runtime/metal/kernels/common/gdn_primitives.h
source/runtime/metal/kernels/common/gguf_embedding_formats.h
source/runtime/metal/kernels/common/gguf_sgmatrix.h
source/runtime/metal/kernels/common/gguf_staged.h
source/runtime/metal/kernels/common/gguf_staged_tile.h
source/runtime/metal/kernels/common/gguf_tile.h
source/runtime/metal/kernels/common/kv_extent.h
source/runtime/metal/kernels/common/kv_paging.h
source/runtime/metal/kernels/common/lane_bindings.h
source/runtime/metal/kernels/common/moe_expert_slab.h
source/runtime/metal/kernels/common/paged_attention_tile.h
source/runtime/metal/kernels/common/paged_store_row.h
source/runtime/metal/kernels/common/q4_mpp_tiles.h
source/runtime/metal/kernels/common/q4_sgmatrix.h
source/runtime/metal/kernels/common/quant_formats.h
source/runtime/metal/kernels/common/rms_inverse.h
source/runtime/metal/kernels/common/sgmatrix.h
source/runtime/metal/kernels/common/split_reduce.h
source/runtime/metal/kernels/common/tq4_attention_tile.h
source/runtime/metal/kernels/common/tq4_paging.h
source/runtime/metal/kernels/common/tq4_store_row.h
source/runtime/metal/kernels/decode/attention_qkv.metal
source/runtime/metal/kernels/decode/draft.metal
source/runtime/metal/kernels/decode/gdn.metal
source/runtime/metal/kernels/decode/linear_gguf_sgmatrix.metal
source/runtime/metal/kernels/decode/linear_q4.metal
source/runtime/metal/kernels/decode/linear_q4_a8.metal
source/runtime/metal/kernels/decode/linear_q4_grid_split.metal
source/runtime/metal/kernels/decode/linear_q4_sgmatrix.metal
source/runtime/metal/kernels/decode/paged_attention.metal
source/runtime/metal/kernels/decode/paged_attention_store.metal
source/runtime/metal/kernels/decode/sampling.metal
source/runtime/metal/kernels/decode/tq4_attention.metal
source/runtime/metal/kernels/decode/tq4_rotated.metal
source/runtime/metal/kernels/decode/tq4_store.metal
source/runtime/metal/kernels/prefill/ane_ffn.metal
source/runtime/metal/kernels/prefill/attention_qkv.metal
source/runtime/metal/kernels/prefill/cpu_up_silu_sum.metal
source/runtime/metal/kernels/prefill/draft_context.metal
source/runtime/metal/kernels/prefill/gdn.metal
source/runtime/metal/kernels/prefill/linear_q4.metal
source/runtime/metal/kernels/prefill/moe.metal
source/runtime/metal/kernels/prefill/normalization.metal
source/runtime/metal/kernels/prefill/paged_attention.metal
source/runtime/metal/kernels/prefill/paged_attention_store.metal
source/runtime/metal/kernels/prefill/tq4_attention.metal
source/runtime/metal/kernels/prefill/tq4_staged.metal
source/runtime/metal/kernels/prefill/tq4_store.metal
source/runtime/metal/kernels/shared/embedding.metal
source/runtime/metal/kernels/shared/gguf_float.metal
source/runtime/metal/kernels/shared/gguf_linear.metal
source/runtime/metal/kernels/shared/gguf_repack.metal
source/runtime/metal/kernels/shared/gguf_rotation.metal
source/runtime/metal/kernels/shared/moe.metal
source/runtime/metal/kernels/shared/moe_gguf.metal
source/runtime/metal/kernels/shared/normalization.metal
source/runtime/metal/kernels/shared/residency.metal
source/runtime/metal/kernels/shared/rope.metal
source/runtime/metal/kernels/shared/row_copy.metal
source/runtime/metal/kernels/shared/vision.metal
source/runtime/model/AffinePlan.hpp
source/runtime/model/AffinePreparation.cpp
source/runtime/model/AffinePreparation.hpp
source/runtime/model/AffineTarget.cpp
source/runtime/model/AffineTarget.hpp
source/runtime/model/Bfloat16.hpp
source/runtime/model/Crc32c.hpp
source/runtime/model/DFlashDraft.cpp
source/runtime/model/DFlashDraft.hpp
source/runtime/model/DraftCheckpoint.cpp
source/runtime/model/DraftCheckpoint.hpp
source/runtime/model/DraftContextPlan.cpp
source/runtime/model/ElasticResidency.hpp
source/runtime/model/GgufFile.cpp
source/runtime/model/GgufFile.hpp
source/runtime/model/GgufImage.cpp
source/runtime/model/GgufImage.hpp
source/runtime/model/GgufImageLayout.hpp
source/runtime/model/GgufPreparation.cpp
source/runtime/model/GgufPreparation.hpp
source/runtime/model/GgufTarget.cpp
source/runtime/model/GgufTarget.hpp
source/runtime/model/HostEmbedding.hpp
source/runtime/model/LookupController.hpp
source/runtime/model/Model.hpp
source/runtime/model/ModelDescriptor.hpp
source/runtime/model/ModelDescriptor.mm
source/runtime/model/ModelFactory.cpp
source/runtime/model/ModelFactory.hpp
source/runtime/model/PhaseResidency.hpp
source/runtime/model/PromptLookup.hpp
source/runtime/model/Qwen3_6Moe.cpp
source/runtime/model/Qwen3_6Moe.hpp
source/runtime/model/Qwen3_8.cpp
source/runtime/model/Qwen3_8.hpp
source/runtime/model/QwenHybridLayout.hpp
source/runtime/model/QwenState.cpp
source/runtime/model/QwenState.hpp
source/runtime/model/QwenTarget.cpp
source/runtime/model/QwenTarget.hpp
source/runtime/model/QwenTargetFiles.hpp
source/runtime/model/QwenTargetLoader.cpp
source/runtime/model/QwenTargetLoader.hpp
source/runtime/model/QwenVision.cpp
source/runtime/model/QwenVision.hpp
source/runtime/model/Runtime.cpp
source/runtime/model/Runtime.hpp
source/runtime/model/RuntimeArenas.cpp
source/runtime/model/RuntimeArenas.hpp
source/runtime/model/SafetensorsCheckpoint.hpp
source/runtime/model/SafetensorsCheckpoint.mm
source/runtime/model/SlotFile.cpp
source/runtime/model/SlotFile.hpp
source/runtime/model/StateLayout.hpp
source/runtime/model/StateTransfer.hpp
source/runtime/model/ThinkingBudget.hpp
source/runtime/model/VisionLoader.cpp
source/runtime/model/VisionLoader.hpp
source/runtime/model/VisionPreparation.cpp
source/runtime/model/VisionPreparation.hpp
source/runtime/model/VisionRows.hpp
source/runtime/model/WeightImages.cpp
source/runtime/model/WeightImages.hpp
source/runtime/model/WeightLayout.hpp
source/runtime/model/WeightMemory.hpp
source/runtime/model/WeightSource.cpp
source/runtime/model/WeightSource.hpp
source/runtime/model/WeightStore.cpp
source/runtime/model/WeightStore.hpp
source/runtime/ops/AneFfn.cpp
source/runtime/ops/AneFfn.hpp
source/runtime/ops/AneFfnCalibration.cpp
source/runtime/ops/AneFfnCalibration.hpp
source/runtime/ops/AneFfnMeasurement.cpp
source/runtime/ops/AneFfnMeasurement.hpp
source/runtime/ops/BufferExtent.hpp
source/runtime/ops/DraftAttention.cpp
source/runtime/ops/DraftAttention.hpp
source/runtime/ops/DraftSelector.cpp
source/runtime/ops/DraftSelector.hpp
source/runtime/ops/Embedding.cpp
source/runtime/ops/Embedding.hpp
source/runtime/ops/ExecutionPlans.cpp
source/runtime/ops/ExecutionPlans.hpp
source/runtime/ops/GDN.cpp
source/runtime/ops/GDN.hpp
source/runtime/ops/HostLinear.cpp
source/runtime/ops/HostLinear.hpp
source/runtime/ops/LaneBindings.hpp
source/runtime/ops/Linear.cpp
source/runtime/ops/Linear.hpp
source/runtime/ops/LinearGguf.cpp
source/runtime/ops/MoE.cpp
source/runtime/ops/MoE.hpp
source/runtime/ops/Normalization.cpp
source/runtime/ops/Normalization.hpp
source/runtime/ops/PageStorage.cpp
source/runtime/ops/PageStorage.hpp
source/runtime/ops/PagedAttention.cpp
source/runtime/ops/PagedAttention.hpp
source/runtime/ops/PagedKv.hpp
source/runtime/ops/RoPE.cpp
source/runtime/ops/RoPE.hpp
source/runtime/ops/RowCopy.cpp
source/runtime/ops/RowCopy.hpp
source/runtime/ops/Sampling.cpp
source/runtime/ops/Sampling.hpp
source/runtime/ops/TQ4ExtentLayout.hpp
source/runtime/ops/TQ4ExtentStorage.cpp
source/runtime/ops/TQ4ExtentStorage.hpp
source/runtime/ops/VerifyShape.hpp
source/runtime/ops/Vision.cpp
source/runtime/ops/Vision.hpp
source/runtime/ops/Weights.hpp
source/server/__init__.py
source/server/ane_model_cache.py
source/server/ane_placement.py
source/server/ane_vision_worker.py
source/server/api_shapes.py
source/server/backend.py
source/server/chat.html
source/server/chat_templates.py
source/server/connections.py
source/server/constraints.py
source/server/crash_trace.py
source/server/diagnostics.py
source/server/document_worker.py
source/server/documents.py
source/server/errors.py
source/server/favicon.svg
source/server/frontend.py
source/server/grouped_vision_profile.py
source/server/http_security.py
source/server/images.py
source/server/json_codec.py
source/server/judgments.py
source/server/latency.py
source/server/lru.py
source/server/metrics.py
source/server/origins.py
source/server/output.py
source/server/preparation_cancel.py
source/server/prepared_vision.py
source/server/protocol.py
source/server/runtime.py
source/server/schema_validation.py
source/server/serve_options.py
source/server/server.py
source/server/thinking.py
source/server/thinking_budget.py
source/server/tokenization.py
source/server/tool_schema.py
source/server/variable_vision_worker.py
source/server/vision_adapter.py
source/server/vision_cache_ready.py
source/server/vision_geometry.py
source/server/vision_routing.py
tests/test_ane_bundle.py
tests/test_branding.py
tests/test_dependency_pins.py
tests/test_grouped_profile.py
tests/test_image_security.py
tests/test_local_launcher.py
tests/test_one_command_model.py
tests/test_one_command_safety.py
tests/test_package_validation.py
tests/test_pip_cli.py
tests/test_preparation_cancel.py
tests/test_runtime_bundle.py
tests/test_thinking_budget.py
tests/test_vision_lifecycle.py
tests/test_vision_provision.py
tests/test_vision_relief.py
tests/test_vision_startup.py
tests/test_worker_packaging.py
tests/thinking_budget_state.cpp
tests/thinking_budget_wire.cpp
tests/fixtures/ane/1080p.png
tests/fixtures/ane/cases.json
tests/fixtures/ane/generate_dashboard.py
tests/fixtures/ane/small.png
vision-tools/CoreAIBridge.swift
vision-tools/ane_model_cache.py
vision-tools/ane_placement.py
vision-tools/attributed_memory.c
vision-tools/coreai_bridge.py
vision-tools/graph.py
vision-tools/grouped_worker.py
vision-tools/memory_audit.py
vision-tools/metered_worker.py
vision-tools/partition.py
vision-tools/provision.py
vision-tools/requirements-export.txt
vision-tools/variable_vision_worker.py
vision-tools/vision_cache_ready.py
vision-tools/vision_geometry.py
vision-tools/vision_graph.py
vision-tools/vision_weights.py
vision-tools/warm_chain.py
vision-tools/worker_meter.py
website/README.md