LICENSE
README.md
pyproject.toml
tests/test_acceleration_contract.py
tests/test_admin_sleep_busy_guard.py
tests/test_affine_jit_default_gate.py
tests/test_affine_moe_pair_decode.py
tests/test_agentic_harness_grading.py
tests/test_agentic_protocol_matrix.py
tests/test_all_local_model_smoke.py
tests/test_answer_pass_families_dsv4_step37.py
tests/test_answer_pass_streaming.py
tests/test_anthropic_adapter.py
tests/test_anthropic_stream_fixes.py
tests/test_api_kwarg_parity.py
tests/test_api_models.py
tests/test_api_surface_contract.py
tests/test_api_surface_parity.py
tests/test_api_utils.py
tests/test_atem_tool_parser.py
tests/test_audio.py
tests/test_audit_fixes.py
tests/test_batched_executor_lifecycle.py
tests/test_batching.py
tests/test_batching_deterministic.py
tests/test_block_cache_namespace_parity.py
tests/test_block_disk_access_only_reconcile.py
tests/test_block_disk_bfloat16.py
tests/test_block_disk_budget_resolution.py
tests/test_block_disk_default.py
tests/test_block_disk_oversized_payload_admission.py
tests/test_block_disk_parent_chain.py
tests/test_block_disk_pending_release_before_eviction.py
tests/test_block_disk_transient_read_keeps_chain.py
tests/test_block_disk_write_admission_timeout.py
tests/test_block_disk_write_fence.py
tests/test_block_disk_write_fence_leak.py
tests/test_block_disk_writer_survives_batch_failure.py
tests/test_bundle_chat_template_audit.py
tests/test_cache_architecture_contract.py
tests/test_cache_bypass.py
tests/test_cache_contract_tokenizer_owner.py
tests/test_cache_extra_key_backends.py
tests/test_cache_fixes.py
tests/test_cache_hierarchy_live_gate.py
tests/test_cache_hit_worker_dequant.py
tests/test_cache_isolation.py
tests/test_cache_list_alignment.py
tests/test_cache_logical_extent.py
tests/test_cache_memory_ceiling_from_os.py
tests/test_cache_record_validator.py
tests/test_cache_segment_boundaries.py
tests/test_cache_types.py
tests/test_cancellation.py
tests/test_chat_template_kwargs.py
tests/test_clean_boundary_assembly.py
tests/test_cli_commands.py
tests/test_cli_panel_settings_parity.py
tests/test_codebook_weight_cache.py
tests/test_config_sources.py
tests/test_context_output_clamp.py
tests/test_continuous_batching.py
tests/test_continuous_batching_extend.py
tests/test_cross_matrix_audit_runner.py
tests/test_cross_matrix_output_counts.py
tests/test_current_regression_suite.py
tests/test_deepseek_r1_reasoning_no_leak.py
tests/test_dflash2_prefix_reuse.py
tests/test_dflash2_runtime.py
tests/test_disk_cache_bf16_lossless.py
tests/test_disk_cache_unit.py
tests/test_disk_prefix_n_minus_1.py
tests/test_doctor_native_mtp.py
tests/test_dots3_language.py
tests/test_dots3_mixed_media.py
tests/test_dots3_padded_restore.py
tests/test_dots3_towers.py
tests/test_dots_tool_parser.py
tests/test_dseries_capability_stamp_routing.py
tests/test_dsml_tool_parser.py
tests/test_dsv4_affine_moe_fastpath.py
tests/test_dsv4_answer_reserve_split.py
tests/test_dsv4_batch_generator_speed.py
tests/test_dsv4_block_size_reconciliation.py
tests/test_dsv4_bundle_integrity.py
tests/test_dsv4_contract_hardening.py
tests/test_dsv4_default_cache_tool_loop_gate.py
tests/test_dsv4_delta_snapshot_batching.py
tests/test_dsv4_delta_transport.py
tests/test_dsv4_equivalence_probe_contract.py
tests/test_dsv4_fastpath_recovery_release.py
tests/test_dsv4_fastpath_v1624_integration.py
tests/test_dsv4_fused_pair_moe.py
tests/test_dsv4_indexed_decode_kernel.py
tests/test_dsv4_paged_cache.py
tests/test_dsv4_performance_suite.py
tests/test_dsv4_pipelined_decode.py
tests/test_dsv4_prefill_snapshot_batching.py
tests/test_dsv4_real_activation_expert_probe.py
tests/test_dsv4_responses_restart_l2_gate.py
tests/test_dsv4_route_mode_code_exactness.py
tests/test_dsv4_safe_fastpaths.py
tests/test_dsv4_terminal_cleanup.py
tests/test_dsv4_thinking_finalizer.py
tests/test_dsv4_thinking_soft_cap.py
tests/test_dsv4_v9_v10_cache_schema_label_mapping.py
tests/test_dsv4_vector_probe.py
tests/test_dup_mlx_preflight_safety.py
tests/test_e2e_live.py
tests/test_embedding_sequence_length.py
tests/test_embeddings.py
tests/test_emoji_comprehensive.py
tests/test_engine_audit.py
tests/test_fail_closed_detection.py
tests/test_fetch_cache_telemetry.py
tests/test_flash_moe_loader.py
tests/test_force_drop_reaches_ordinary_kv.py
tests/test_full_release_objective_checklist.py
tests/test_gemma3_27b_comprehensive.py
tests/test_gemma3_tool_parser.py
tests/test_gemma4_12b_artifact_contract.py
tests/test_gemma4_12b_live_runtime_audit.py
tests/test_gemma4_12b_speed_gate.py
tests/test_gemma4_image_budget.py
tests/test_gemma4_reasoning_no_leak.py
tests/test_gemma4_tool_parser.py
tests/test_gemma_leak.py
tests/test_gemma_qat_native_mxfp4_inventory_gate.py
tests/test_generation_defaults_contract.py
tests/test_generation_prompt_cache_key.py
tests/test_generator_insert_signature_parity.py
tests/test_glm47_tool_parser.py
tests/test_glm5_next_runtime.py
tests/test_global_disk_cache_budget.py
tests/test_gptoss_reasoning.py
tests/test_granite_tool_parser.py
tests/test_health_endpoint.py
tests/test_health_live_overlay_helpers.py
tests/test_hunyuan_tool_parser.py
tests/test_hy3_native_mtp.py
tests/test_hybrid_batching.py
tests/test_hybrid_chunked_prefill_equivalence.py
tests/test_hybrid_kv_estimate.py
tests/test_hybrid_live_tq_kv.py
tests/test_hybrid_prefill_valve.py
tests/test_hybrid_prefix_cache.py
tests/test_hybrid_prefix_cache_clone.py
tests/test_hybrid_prefix_promotion_gate.py
tests/test_hybrid_ssm_companion_regressions.py
tests/test_image_api.py
tests/test_image_gen.py
tests/test_image_gen_engine.py
tests/test_image_stays_in_its_own_message.py
tests/test_impl_campaign_20260710.py
tests/test_inbound_request_field_logging.py
tests/test_inkling_reasoning_parser.py
tests/test_installed_app_runtime_parity_audit.py
tests/test_issue175_177_installed_runtime_audit.py
tests/test_issue175_177_live_runtime_audit.py
tests/test_issue175_179_release_boundary_audit.py
tests/test_issue175_admin_sleep_probe.py
tests/test_issue179_minimax_k_model_manifest.py
tests/test_issue179_minimax_k_root_cause_audit.py
tests/test_issue179_reporter_parity_metadata.py
tests/test_issue179_responses_cancel_probe.py
tests/test_issue181_183_runtime_audit.py
tests/test_jang_affine_storage.py
tests/test_jang_loader.py
tests/test_jang_loader_quantize_predicate_guard.py
tests/test_jang_model_compat_pr155.py
tests/test_jang_vlm_jangtq_envelope.py
tests/test_jit_toggle.py
tests/test_kimi_k25_mla_patch.py
tests/test_kv_byte_estimator_recurses.py
tests/test_kv_quant.py
tests/test_kv_quant_default_policy.py
tests/test_kv_quant_status_truthfulness.py
tests/test_laguna_explicit_thinking_budget.py
tests/test_laguna_loader.py
tests/test_lfm2_vl_compat.py
tests/test_lifecycle_load_progress.py
tests/test_llm.py
tests/test_local_generation_metadata_audit.py
tests/test_low_ram_cache_advice.py
tests/test_max_output_context_contract.py
tests/test_mcp_policy.py
tests/test_mcp_policy_contract.py
tests/test_mcp_sdk_compat.py
tests/test_mcp_security.py
tests/test_medgemma_comprehensive.py
tests/test_media_chunked_prefill.py
tests/test_media_clean_boundary_alignment.py
tests/test_media_history_prefix_cap.py
tests/test_memory_cache.py
tests/test_memory_cache_lock_inversion.py
tests/test_memory_limits.py
tests/test_metal_buffer_count_guard.py
tests/test_metal_headroom_output_guard.py
tests/test_mimo_completion_chat_rail.py
tests/test_mimo_v2_audio_template.py
tests/test_mimo_v2_cache_vs_nocache_next_token.py
tests/test_mimo_v2_current_audit.py
tests/test_mimo_v2_exactness_variant_probe.py
tests/test_mimo_v2_local_bundle_metadata_contract.py
tests/test_mimo_v2_media_capability_gate.py
tests/test_mimo_v2_media_runtime.py
tests/test_mimo_v2_mllm_runtime_registration.py
tests/test_mimo_v2_no_source_exactness_classifier.py
tests/test_mimo_v2_rotating_cache_patch.py
tests/test_mimo_v2_source_vs_quant_probe.py
tests/test_minimax_m3_cache_paths.py
tests/test_minimax_m3_jang_loader_quant.py
tests/test_minimax_m3_tool_parser.py
tests/test_mistral3_loader.py
tests/test_mixed_swa_stored_kv_exact.py
tests/test_mllm.py
tests/test_mllm_cache.py
tests/test_mllm_continuous_batching.py
tests/test_mllm_lce_publication.py
tests/test_mllm_message_serialization.py
tests/test_mllm_native_mtp_tq_snapshot.py
tests/test_mllm_processor_stream_contracts.py
tests/test_mllm_sampling_contract.py
tests/test_mllm_scheduler_cache.py
tests/test_mllm_scheduler_stability.py
tests/test_mllm_stream_lifecycle.py
tests/test_mllm_streaming_finalization.py
tests/test_mllm_tool_replay.py
tests/test_mlx0322_qwen3_vl_repeat_compat.py
tests/test_mlx_memory_cleanup.py
tests/test_mlx_memory_dtype.py
tests/test_mlx_vlm_patch_targets_still_exist.py
tests/test_model_artifact_format_contract.py
tests/test_model_config_registry.py
tests/test_model_family_detection_contract.py
tests/test_model_inspector.py
tests/test_model_name.py
tests/test_model_registry.py
tests/test_mtp_depth_verify_cycle.py
tests/test_mtp_telemetry_edge_cases.py
tests/test_multi_eos.py
tests/test_multimodal_routing.py
tests/test_muse_glimmer_marker_leak.py
tests/test_muse_glimmer_port.py
tests/test_muse_glimmer_processor.py
tests/test_muse_glimmer_reasoning_parser.py
tests/test_muse_glimmer_registration.py
tests/test_muse_glimmer_registry.py
tests/test_n2_chat_cache_gate.py
tests/test_nanbeige_runtime_contract.py
tests/test_native_mtp_adaptive_raise.py
tests/test_native_mtp_adaptive_value.py
tests/test_native_mtp_ar_fallback_sample.py
tests/test_native_mtp_autodetect.py
tests/test_native_mtp_bench_harness.py
tests/test_native_mtp_contract.py
tests/test_native_mtp_cycle_fence_depth.py
tests/test_native_mtp_decision_bundle.py
tests/test_native_mtp_examples.py
tests/test_native_mtp_policy_suite.py
tests/test_native_mtp_profile.py
tests/test_native_mtp_prompt_priming.py
tests/test_native_mtp_research_sampling.py
tests/test_native_mtp_silent_skip.py
tests/test_native_mtp_stochastic_accept.py
tests/test_native_mtp_verify_qmm.py
tests/test_native_tool_format.py
tests/test_nemotron_h_v2_model_type_gates.py
tests/test_nonstream_prefill_grace.py
tests/test_objective_proof_digest.py
tests/test_offpath_disk_writes.py
tests/test_ollama_adapter.py
tests/test_ollama_durations.py
tests/test_ollama_reasoning_parity.py
tests/test_ollama_show_tools_capability.py
tests/test_omni_multimodal.py
tests/test_openai_token_penalties.py
tests/test_openpangu_tool_parser.py
tests/test_openpangu_v2.py
tests/test_optimizations.py
tests/test_output_budget_cap.py
tests/test_output_budget_enforcement.py
tests/test_packaged_integrity_contract.py
tests/test_paged_byte_budget.py
tests/test_paged_byte_budget_gate.py
tests/test_paged_cache.py
tests/test_paged_cache_benefits.py
tests/test_paged_cache_hash_collisions.py
tests/test_paged_cache_real_inference.py
tests/test_paged_cache_real_model.py
tests/test_paged_cache_unit.py
tests/test_paged_clear_and_free_guards.py
tests/test_paged_default_family_parity.py
tests/test_paged_hit_touches_l2_access.py
tests/test_paged_keep_resident_budget.py
tests/test_paged_recycler_clears_stale_payload.py
tests/test_panel_cli_flag_contract.py
tests/test_param_parity_fixes.py
tests/test_parser_boundary_variation_probe_contract.py
tests/test_parser_registry_contract.py
tests/test_payload_byte_walker_contract.py
tests/test_perf_prefill_loop.py
tests/test_perf_safe_wins.py
tests/test_perf_sampler_async.py
tests/test_platform.py
tests/test_prefill_admission.py
tests/test_prefix_cache.py
tests/test_prefix_cache_byte_budget_both_paths.py
tests/test_prefix_cache_evicted_l2_dedup.py
tests/test_prefix_cache_key_single_source.py
tests/test_prefix_index_hash_chain.py
tests/test_public_app_issue_audit.py
tests/test_quant_status_generic_mlx_stamp.py
tests/test_quantized_kv_dtype_roundtrip.py
tests/test_quantized_projection_group.py
tests/test_qwen35_family_vendor_overlay.py
tests/test_qwen35_projection_groups.py
tests/test_qwen35_vl_mtp_sanitize.py
tests/test_qwen36_tp4_gateway_probe.py
tests/test_qwen3_answer_pass_policy.py
tests/test_qwen3_coder_tool_parser_name.py
tests/test_qwen4_exp_runtime.py
tests/test_qwen_forward_path_ab.py
tests/test_qwen_gated_delta_subprofile.py
tests/test_qwen_layer_profile.py
tests/test_real_ui_dsv4_memory_preflight.py
tests/test_reasoning_marker_chunk_boundaries.py
tests/test_reasoning_modes.py
tests/test_reasoning_parser.py
tests/test_reasoning_parser_explicit_disable.py
tests/test_reasoning_parser_whitespace_fidelity.py
tests/test_reasoning_template_contract.py
tests/test_reasoning_tool_interaction.py
tests/test_reconstruct_memo_headroom.py
tests/test_release_automation.py
tests/test_release_gate_python_app.py
tests/test_release_regression_manifest.py
tests/test_release_surface_contract.py
tests/test_remote_max2_dsv4_exactness_guard.py
tests/test_remote_max2_dsv4_readiness.py
tests/test_remote_max2_dsv4_real_ui_guard.py
tests/test_remote_max2_dsv4_release_proof_guard.py
tests/test_request.py
tests/test_request_cancellation.py
tests/test_request_progress_monotonic.py
tests/test_reranker_endpoint.py
tests/test_responses_history.py
tests/test_responses_multimodal_history.py
tests/test_responses_replay_call_id_stability.py
tests/test_runtime_cache_fingerprint_source_id.py
tests/test_runtime_memory_stress_probe.py
tests/test_sampling.py
tests/test_scheduler_cache_init_cleanup.py
tests/test_scheduler_cache_truth.py
tests/test_scheduler_insert_capability.py
tests/test_scheduler_repetition_context.py
tests/test_scoped_release_preflight_16.py
tests/test_scoped_release_preflight_17.py
tests/test_scoped_release_preflight_20.py
tests/test_scoped_release_preflight_66.py
tests/test_server.py
tests/test_server_contentless_turns.py
tests/test_server_module_entry_delegates_to_cli.py
tests/test_simple_engine.py
tests/test_simple_engine_prompt_accounting.py
tests/test_single_active_batch_generator.py
tests/test_single_sequence_cache_merge.py
tests/test_slow_family_timeout_reaches_streaming.py
tests/test_smelt_loader.py
tests/test_sparse_index_score_decode.py
tests/test_special_tag_dialect.py
tests/test_speculative.py
tests/test_ssm_companion_cache.py
tests/test_stamped_effort_policy.py
tests/test_stamped_reasoning_efforts.py
tests/test_step37_crash_falsification_contract.py
tests/test_step37_mlx_vlm_runtime.py
tests/test_step37_vlm_runtime_audit.py
tests/test_step3p5_tool_parser.py
tests/test_step3p7_mllm_detection_guard.py
tests/test_store_writes_logical_extent_only.py
tests/test_stream_cancel_abort.py
tests/test_stream_keepalive_timeout.py
tests/test_stream_terminal_finish_guard.py
tests/test_streaming_adapter_failures.py
tests/test_streaming_detokenizer.py
tests/test_streaming_json_encoder.py
tests/test_streaming_latency.py
tests/test_streaming_reasoning.py
tests/test_structured_output.py
tests/test_structured_output_repair_report.py
tests/test_terminal_dispatch_before_cache_cleanup.py
tests/test_thermal_state.py
tests/test_think_xml_parser.py
tests/test_thinking_template_render.py
tests/test_tokenizer_chat_template_include.py
tests/test_tool_call_contract.py
tests/test_tool_fallback_injection.py
tests/test_tool_format.py
tests/test_tool_marker_registry_contract.py
tests/test_tool_parser_argument_whitespace.py
tests/test_tool_parsers.py
tests/test_tool_prompt_fallback.py
tests/test_tq_disk_cache.py
tests/test_tq_e2e_live.py
tests/test_tq_layer_batch_decode.py
tests/test_tq_paged_block_cache.py
tests/test_turboquant_cache_contract.py
tests/test_turboquant_prefix_cache_clone.py
tests/test_vl_media_cache_contract.py
tests/test_vl_video_regression.py
tests/test_vlm_dtype_promotion_guard.py
tests/test_vlm_image_prefill_message.py
tests/test_wake_serialization.py
tests/test_warm_prefix_no_double_release.py
tests/test_xml_function_tool_parser.py
tests/test_zaya_conversion_routing.py
tests/test_zaya_prefix_reuse_gate.py
tests/test_zaya_runtime.py
tests/test_zero_cache_memory_is_not_unbounded.py
tests/test_zero_centered_norm_shift.py
vmlx.egg-info/PKG-INFO
vmlx.egg-info/SOURCES.txt
vmlx.egg-info/dependency_links.txt
vmlx.egg-info/entry_points.txt
vmlx.egg-info/requires.txt
vmlx.egg-info/top_level.txt
vmlx_engine/__init__.py
vmlx_engine/acceleration_contract.py
vmlx_engine/attention.py
vmlx_engine/benchmark.py
vmlx_engine/block_disk_store.py
vmlx_engine/cache_key.py
vmlx_engine/cache_record_validator.py
vmlx_engine/cli.py
vmlx_engine/context_limits.py
vmlx_engine/dflash2_runtime.py
vmlx_engine/disk_cache.py
vmlx_engine/embedding.py
vmlx_engine/engine_core.py
vmlx_engine/errors.py
vmlx_engine/flash_moe_config.py
vmlx_engine/global_disk_cache_budget.py
vmlx_engine/gradio_app.py
vmlx_engine/image_gen.py
vmlx_engine/load_progress.py
vmlx_engine/logprobs.py
vmlx_engine/memory_cache.py
vmlx_engine/mllm_batch_generator.py
vmlx_engine/mllm_cache.py
vmlx_engine/mllm_scheduler.py
vmlx_engine/mlx_memory.py
vmlx_engine/mlx_platform.py
vmlx_engine/model_config_registry.py
vmlx_engine/model_configs.py
vmlx_engine/model_registry.py
vmlx_engine/model_runner.py
vmlx_engine/multimodal_processor.py
vmlx_engine/native_mtp.py
vmlx_engine/native_mtp_acceptance.py
vmlx_engine/native_mtp_adaptive.py
vmlx_engine/native_mtp_cache_telemetry.py
vmlx_engine/native_mtp_policy_suite.py
vmlx_engine/native_mtp_profile.py
vmlx_engine/native_mtp_prompt_priming.py
vmlx_engine/native_mtp_research.py
vmlx_engine/omni_multimodal.py
vmlx_engine/optimizations.py
vmlx_engine/output_collector.py
vmlx_engine/paged_cache.py
vmlx_engine/plugin.py
vmlx_engine/prefix_cache.py
vmlx_engine/prompt_lookup.py
vmlx_engine/request.py
vmlx_engine/reranker.py
vmlx_engine/sampling.py
vmlx_engine/scheduler.py
vmlx_engine/server.py
vmlx_engine/special_tag_dialect.py
vmlx_engine/speculative.py
vmlx_engine/state_machine.py
vmlx_engine/tq_disk_store.py
vmlx_engine/vision_embedding_cache.py
vmlx_engine/worker.py
vmlx_engine/api/__init__.py
vmlx_engine/api/anthropic_adapter.py
vmlx_engine/api/models.py
vmlx_engine/api/ollama_adapter.py
vmlx_engine/api/streaming.py
vmlx_engine/api/tool_calling.py
vmlx_engine/api/utils.py
vmlx_engine/audio/__init__.py
vmlx_engine/audio/processor.py
vmlx_engine/audio/stt.py
vmlx_engine/audio/tts.py
vmlx_engine/cache/__init__.py
vmlx_engine/cache/byte_estimators.py
vmlx_engine/cache/codebook_cache.py
vmlx_engine/chat_templates/gemma4.jinja
vmlx_engine/commands/__init__.py
vmlx_engine/commands/convert.py
vmlx_engine/commands/doctor.py
vmlx_engine/commands/info.py
vmlx_engine/commands/list.py
vmlx_engine/config/__init__.py
vmlx_engine/config/defaults.yaml
vmlx_engine/config/manager.py
vmlx_engine/config/models.py
vmlx_engine/config/sources/__init__.py
vmlx_engine/config/sources/cli_source.py
vmlx_engine/config/sources/env_source.py
vmlx_engine/config/sources/yaml_source.py
vmlx_engine/distributed/__init__.py
vmlx_engine/distributed/cli.py
vmlx_engine/distributed/coordinator.py
vmlx_engine/distributed/discovery.py
vmlx_engine/distributed/engine.py
vmlx_engine/distributed/layer_assign.py
vmlx_engine/distributed/mesh_election.py
vmlx_engine/distributed/mesh_manager.py
vmlx_engine/distributed/mesh_node.py
vmlx_engine/distributed/mesh_topology.py
vmlx_engine/distributed/protocol.py
vmlx_engine/distributed/tensor_parallel.py
vmlx_engine/distributed/worker.py
vmlx_engine/engine/__init__.py
vmlx_engine/engine/base.py
vmlx_engine/engine/batched.py
vmlx_engine/engine/simple.py
vmlx_engine/loaders/__init__.py
vmlx_engine/loaders/dsv4_chat_encoder.py
vmlx_engine/loaders/encoding_dsv4_0731.py
vmlx_engine/loaders/load_jangtq.py
vmlx_engine/loaders/load_jangtq_dsv4.py
vmlx_engine/loaders/load_jangtq_kimi_vlm.py
vmlx_engine/loaders/load_jangtq_vlm.py
vmlx_engine/loaders/load_laguna.py
vmlx_engine/loaders/load_mistral3.py
vmlx_engine/loaders/load_zaya.py
vmlx_engine/mcp/__init__.py
vmlx_engine/mcp/client.py
vmlx_engine/mcp/config.py
vmlx_engine/mcp/executor.py
vmlx_engine/mcp/manager.py
vmlx_engine/mcp/security.py
vmlx_engine/mcp/tools.py
vmlx_engine/mcp/types.py
vmlx_engine/metal/__init__.py
vmlx_engine/metal/affine_moe_decode.metal
vmlx_engine/metal/affine_moe_decode.py
vmlx_engine/metal/affine_moe_pair_decode.py
vmlx_engine/metal/codebook_matvec.metal
vmlx_engine/metal/codebook_moe.metal
vmlx_engine/metal/fused_pair_moe_decode.py
vmlx_engine/metal/gated_rmsnorm_decode.py
vmlx_engine/metal/gdn_conv_decode.py
vmlx_engine/metal/glm5_hc_place_decode.py
vmlx_engine/metal/glm5_mhc_decode.py
vmlx_engine/metal/kda_conv_decode.py
vmlx_engine/metal/kda_step_decode.py
vmlx_engine/metal/kernel_manager.py
vmlx_engine/metal/native_mtp_verify_qmm.py
vmlx_engine/metal/ple_conv_decode.py
vmlx_engine/metal/quantized_projection_group.py
vmlx_engine/metal/qwen35_gdn_gate_terms.py
vmlx_engine/metal/qwen4_affine_moe_decode.py
vmlx_engine/metal/sparse_index_score_decode.py
vmlx_engine/models/__init__.py
vmlx_engine/models/_gemma4_text_upstream.py
vmlx_engine/models/_gemma4_upstream.py
vmlx_engine/models/codebook.py
vmlx_engine/models/codebook_expert_loader.py
vmlx_engine/models/codebook_moe_integration.py
vmlx_engine/models/dots3_note_register.py
vmlx_engine/models/dsv4_lm_head_fastpath.py
vmlx_engine/models/dsv4_rope_cache.py
vmlx_engine/models/flash_moe_integration.py
vmlx_engine/models/gemma4_native_register.py
vmlx_engine/models/gemma4_unified_register.py
vmlx_engine/models/llm.py
vmlx_engine/models/mllm.py
vmlx_engine/models/muse_glimmer_register.py
vmlx_engine/models/step3p7_mlx_vlm.py
vmlx_engine/models/zaya.py
vmlx_engine/models/zaya1_vl.py
vmlx_engine/models/dots3_note/__init__.py
vmlx_engine/models/dots3_note/audio.py
vmlx_engine/models/dots3_note/config.py
vmlx_engine/models/dots3_note/dots3_note.py
vmlx_engine/models/dots3_note/language.py
vmlx_engine/models/dots3_note/processor.py
vmlx_engine/models/dots3_note/vision.py
vmlx_engine/models/gemma4_unified/__init__.py
vmlx_engine/models/gemma4_unified/config.py
vmlx_engine/models/gemma4_unified/gemma4_unified.py
vmlx_engine/models/gemma4_unified/processing_gemma4_unified.py
vmlx_engine/models/glm5_next/__init__.py
vmlx_engine/models/glm5_next/glm5_next.py
vmlx_engine/models/glm5_next/kda.py
vmlx_engine/models/glm5_next/register.py
vmlx_engine/models/minimax_m3/__init__.py
vmlx_engine/models/minimax_m3/cache.py
vmlx_engine/models/minimax_m3/eagle3_draft.py
vmlx_engine/models/minimax_m3/m3_affine2_switch.py
vmlx_engine/models/minimax_m3/m3_vl_preprocess.py
vmlx_engine/models/minimax_m3/minimax_m3.py
vmlx_engine/models/minimax_m3/minimax_m3_register.py
vmlx_engine/models/minimax_m3/minimax_m3_vl.py
vmlx_engine/models/minimax_m3/runtime.py
vmlx_engine/models/muse_glimmer/__init__.py
vmlx_engine/models/muse_glimmer/config.py
vmlx_engine/models/muse_glimmer/language.py
vmlx_engine/models/muse_glimmer/muse_glimmer.py
vmlx_engine/models/muse_glimmer/processor.py
vmlx_engine/models/muse_glimmer/vision.py
vmlx_engine/models/openpangu_v2/__init__.py
vmlx_engine/models/openpangu_v2/cache.py
vmlx_engine/models/openpangu_v2/openpangu_v2.py
vmlx_engine/models/openpangu_v2/register.py
vmlx_engine/models/qwen3_5_family/__init__.py
vmlx_engine/models/qwen3_5_family/register.py
vmlx_engine/models/qwen3_5_family/qwen3_5/__init__.py
vmlx_engine/models/qwen3_5_family/qwen3_5/config.py
vmlx_engine/models/qwen3_5_family/qwen3_5/language.py
vmlx_engine/models/qwen3_5_family/qwen3_5/qwen3_5.py
vmlx_engine/models/qwen3_5_family/qwen3_5/vision.py
vmlx_engine/models/qwen3_5_family/qwen3_5_moe/__init__.py
vmlx_engine/models/qwen3_5_family/qwen3_5_moe/config.py
vmlx_engine/models/qwen3_5_family/qwen3_5_moe/language.py
vmlx_engine/models/qwen3_5_family/qwen3_5_moe/qwen3_5_moe.py
vmlx_engine/models/qwen3_5_family/qwen3_5_moe/vision.py
vmlx_engine/models/qwen4_exp/__init__.py
vmlx_engine/models/qwen4_exp/config.py
vmlx_engine/models/qwen4_exp/language.py
vmlx_engine/models/qwen4_exp/loader.py
vmlx_engine/models/qwen4_exp/ngram.py
vmlx_engine/models/qwen4_exp/qwen4_exp.py
vmlx_engine/models/qwen4_exp/register.py
vmlx_engine/models/qwen4_exp/table_reader.py
vmlx_engine/models/qwen4_exp/vision.py
vmlx_engine/native_mtp_examples/README.md
vmlx_engine/native_mtp_examples/__init__.py
vmlx_engine/native_mtp_examples/env_flag_matrix.py
vmlx_engine/native_mtp_examples/generate_server_command.py
vmlx_engine/native_mtp_examples/inspect_mtp_metadata.py
vmlx_engine/native_mtp_examples/mtp_runtime_common.py
vmlx_engine/native_mtp_examples/parse_mtp_logs.py
vmlx_engine/native_mtp_examples/server_smoke.py
vmlx_engine/patches/__init__.py
vmlx_engine/patches/mlx_lm_mtp/__init__.py
vmlx_engine/patches/mlx_lm_mtp/batch_generator.py
vmlx_engine/patches/mlx_lm_mtp/cache_rollback.py
vmlx_engine/patches/mlx_lm_mtp/deepseek_v4_model.py
vmlx_engine/patches/mlx_lm_mtp/hy_v3_model.py
vmlx_engine/patches/mlx_lm_mtp/qwen35_model.py
vmlx_engine/patches/mlx_vlm_mtp/__init__.py
vmlx_engine/patches/mlx_vlm_mtp/qwen35_vl.py
vmlx_engine/reasoning/__init__.py
vmlx_engine/reasoning/base.py
vmlx_engine/reasoning/deepseek_r1_parser.py
vmlx_engine/reasoning/gemma4_parser.py
vmlx_engine/reasoning/gptoss_parser.py
vmlx_engine/reasoning/inkling_parser.py
vmlx_engine/reasoning/minimax_m2_parser.py
vmlx_engine/reasoning/minimax_m3_parser.py
vmlx_engine/reasoning/mistral_parser.py
vmlx_engine/reasoning/muse_glimmer_parser.py
vmlx_engine/reasoning/qwen3_parser.py
vmlx_engine/reasoning/think_parser.py
vmlx_engine/reasoning/think_xml_parser.py
vmlx_engine/runtime_patches/__init__.py
vmlx_engine/runtime_patches/deepseek_v3_patched.py
vmlx_engine/runtime_patches/deepseek_v4_register.py
vmlx_engine/runtime_patches/gemma4_processing.py
vmlx_engine/runtime_patches/gemma4_vision.py
vmlx_engine/runtime_patches/kimi_k25_mla.py
vmlx_engine/tests/test_codebook_inference.py
vmlx_engine/tests/test_codebook_kernels.py
vmlx_engine/tool_parsers/__init__.py
vmlx_engine/tool_parsers/abstract_tool_parser.py
vmlx_engine/tool_parsers/atem_tool_parser.py
vmlx_engine/tool_parsers/auto_tool_parser.py
vmlx_engine/tool_parsers/deepseek_tool_parser.py
vmlx_engine/tool_parsers/dots_tool_parser.py
vmlx_engine/tool_parsers/dsml_tool_parser.py
vmlx_engine/tool_parsers/functionary_tool_parser.py
vmlx_engine/tool_parsers/gemma3_tool_parser.py
vmlx_engine/tool_parsers/gemma4_tool_parser.py
vmlx_engine/tool_parsers/glm47_tool_parser.py
vmlx_engine/tool_parsers/granite_tool_parser.py
vmlx_engine/tool_parsers/hermes_tool_parser.py
vmlx_engine/tool_parsers/hunyuan_tool_parser.py
vmlx_engine/tool_parsers/kimi_tool_parser.py
vmlx_engine/tool_parsers/lfm2_tool_parser.py
vmlx_engine/tool_parsers/llama_tool_parser.py
vmlx_engine/tool_parsers/minimax_m3_tool_parser.py
vmlx_engine/tool_parsers/minimax_tool_parser.py
vmlx_engine/tool_parsers/mistral_tool_parser.py
vmlx_engine/tool_parsers/nemotron_tool_parser.py
vmlx_engine/tool_parsers/openpangu_tool_parser.py
vmlx_engine/tool_parsers/qwen_tool_parser.py
vmlx_engine/tool_parsers/step3p5_tool_parser.py
vmlx_engine/tool_parsers/xlam_tool_parser.py
vmlx_engine/tool_parsers/xml_function_tool_parser.py
vmlx_engine/tool_parsers/zaya_tool_parser.py
vmlx_engine/utils/__init__.py
vmlx_engine/utils/cache_extent.py
vmlx_engine/utils/cache_types.py
vmlx_engine/utils/chat_template_kwargs.py
vmlx_engine/utils/chat_templates.py
vmlx_engine/utils/dsv4_batch_generator.py
vmlx_engine/utils/flash_moe_loader.py
vmlx_engine/utils/head_dim_detection.py
vmlx_engine/utils/hybrid_tq_cache.py
vmlx_engine/utils/jang_affine_storage.py
vmlx_engine/utils/jang_loader.py
vmlx_engine/utils/mamba_cache.py
vmlx_engine/utils/memory_limits.py
vmlx_engine/utils/mlx_vlm_compat.py
vmlx_engine/utils/model_inspector.py
vmlx_engine/utils/multi_eos.py
vmlx_engine/utils/nanbeige_runtime.py
vmlx_engine/utils/nemotron_latent_moe.py
vmlx_engine/utils/prefill_admission.py
vmlx_engine/utils/prefix_hit.py
vmlx_engine/utils/quant_shape_inference.py
vmlx_engine/utils/single_batch_generator.py
vmlx_engine/utils/smelt_loader.py
vmlx_engine/utils/ssm_companion_cache.py
vmlx_engine/utils/ssm_companion_disk_store.py
vmlx_engine/utils/thermal.py
vmlx_engine/utils/token_logits_processors.py
vmlx_engine/utils/tokenizer.py
vmlx_engine/utils/turboquant_config.py
vmlx_engine/utils/zero_centered_norms.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/__init__.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/adaptor_base.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/adaptor_generic.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/adaptor_mlp.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/adaptor_registry.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/cls_token.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/common.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/dinov2_arch.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/dual_hybrid_vit.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/enable_cpe_support.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/enable_spectral_reparam.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/eradio_model.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/extra_models.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/extra_timm_models.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/feature_normalizer.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/forward_intermediates.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/hf_model.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/input_conditioner.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/open_clip_adaptor.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/radio_model.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/vit_patch_generator.py
vmlx_engine/vendor/transformers_modules/nvidia/C_hyphen_RADIOv2_hyphen_H/0d8f4c18c877166eda07ddae1386bcad256b7a6a/vitdet.py
vmlx_engine/vlm/__init__.py
vmlx_engine/vlm/generate_vl.py