https://posttrainllm.comhttps://posttrainllm.com/artifactshttps://posttrainllm.com/artifacts/ane-m8-coreml-chainhttps://posttrainllm.com/artifacts/browser-webgpu-speeduphttps://posttrainllm.com/artifacts/factory-run-schema-v1https://posttrainllm.com/artifacts/gallery-int4-browser-modelshttps://posttrainllm.com/artifacts/hf-specialist-model-archive-v1https://posttrainllm.com/artifacts/huge-decode-throughputhttps://posttrainllm.com/artifacts/memory64-browser-behemothhttps://posttrainllm.com/artifacts/qwen3-4b-file-ops-distilledhttps://posttrainllm.com/artifacts/qwen06-sql-routed-v1https://posttrainllm.com/devloghttps://posttrainllm.com/docshttps://posttrainllm.com/docs/agent_runtimehttps://posttrainllm.com/docs/archive/annotated_transcripthttps://posttrainllm.com/docs/archive/evaluationhttps://posttrainllm.com/docs/archive/learning_roadmaphttps://posttrainllm.com/docs/archive/lessonshttps://posttrainllm.com/docs/archive/parked_multi_modelhttps://posttrainllm.com/docs/archive/phase_9_10_statushttps://posttrainllm.com/docs/archive/recent_researchhttps://posttrainllm.com/docs/archive/shared_vs_nativehttps://posttrainllm.com/docs/archive/statushttps://posttrainllm.com/docs/archive/watch_the_model_thinkhttps://posttrainllm.com/docs/artifacts/github-corpus-parityhttps://posttrainllm.com/docs/async_tool_dispatchhttps://posttrainllm.com/docs/attempt-ledgerhttps://posttrainllm.com/docs/audit_2026https://posttrainllm.com/docs/backloghttps://posttrainllm.com/docs/benchmark_first_runhttps://posttrainllm.com/docs/benchmark_harness_designhttps://posttrainllm.com/docs/bpe_browser_scoringhttps://posttrainllm.com/docs/browser_noteshttps://posttrainllm.com/docs/capability_matrixhttps://posttrainllm.com/docs/CITATIONShttps://posttrainllm.com/docs/cold_start_resultshttps://posttrainllm.com/docs/constrained_generationhttps://posttrainllm.com/docs/continue_providerhttps://posttrainllm.com/docs/cpu_speedup_resultshttps://posttrainllm.com/docs/cpu_utilization_researchhttps://posttrainllm.com/docs/data_inventoryhttps://posttrainllm.com/docs/data_perfhttps://posttrainllm.com/docs/dataset-inventoryhttps://posttrainllm.com/docs/decision_loghttps://posttrainllm.com/docs/deployhttps://posttrainllm.com/docs/determinismhttps://posttrainllm.com/docs/distillationhttps://posttrainllm.com/docs/doc-statushttps://posttrainllm.com/docs/docs-quality-audithttps://posttrainllm.com/docs/DRILLDOWNhttps://posttrainllm.com/docs/evolution_strategieshttps://posttrainllm.com/docs/exactness-completion-audithttps://posttrainllm.com/docs/external-products-reviewedhttps://posttrainllm.com/docs/fa2_backward_noteshttps://posttrainllm.com/docs/fa2_forward_noteshttps://posttrainllm.com/docs/factory/batch-posttraininghttps://posttrainllm.com/docs/factory/case-study-templatehttps://posttrainllm.com/docs/factory/enforcementhttps://posttrainllm.com/docs/factory/eval-protocolhttps://posttrainllm.com/docs/factory/huggingface-artifact-storagehttps://posttrainllm.com/docs/factory/lora-geometryhttps://posttrainllm.com/docs/factory/overviewhttps://posttrainllm.com/docs/factory/packaginghttps://posttrainllm.com/docs/factory/post-training-factoryhttps://posttrainllm.com/docs/factory/public-artifactshttps://posttrainllm.com/docs/factory/READMEhttps://posttrainllm.com/docs/factory/reportshttps://posttrainllm.com/docs/factory/run-schemahttps://posttrainllm.com/docs/feature_audit_2026_05_31https://posttrainllm.com/docs/feature_ideashttps://posttrainllm.com/docs/gallery_v2_planhttps://posttrainllm.com/docs/galore_and_stabilityhttps://posttrainllm.com/docs/github_data_integrationhttps://posttrainllm.com/docs/gradient_checkpointing_resultshttps://posttrainllm.com/docs/GRPO_CLARIFYhttps://posttrainllm.com/docs/hf_datasets_integrationhttps://posttrainllm.com/docs/history-coverage-audithttps://posttrainllm.com/docs/industry_learning_roadmaphttps://posttrainllm.com/docs/interpretabilityhttps://posttrainllm.com/docs/kv_cache_optimizationhttps://posttrainllm.com/docs/leaderboardhttps://posttrainllm.com/docs/learnhttps://posttrainllm.com/docs/learn/advanced-llm-inferencehttps://posttrainllm.com/docs/learn/advanced-llm-traininghttps://posttrainllm.com/docs/learn/advanced-ml-systems-evalhttps://posttrainllm.com/docs/learn/agent-context-hierarchyhttps://posttrainllm.com/docs/learn/ane-research/dossierhttps://posttrainllm.com/docs/learn/ane-research/m6-findingshttps://posttrainllm.com/docs/learn/app-intents-comparisonhttps://posttrainllm.com/docs/learn/apple-on-device-foundation-modelshttps://posttrainllm.com/docs/learn/archive/session-01-neural-net-basics-densehttps://posttrainllm.com/docs/learn/castform-rl-finetunehttps://posttrainllm.com/docs/learn/competitive-landscapehttps://posttrainllm.com/docs/learn/coverage-maphttps://posttrainllm.com/docs/learn/curriculumhttps://posttrainllm.com/docs/learn/diversity-driven-small-model-reasoninghttps://posttrainllm.com/docs/learn/essential-vs-optimizationhttps://posttrainllm.com/docs/learn/eval-matrix-2026-06-08https://posttrainllm.com/docs/learn/eval-methodology-2026-06-08https://posttrainllm.com/docs/learn/external-referenceshttps://posttrainllm.com/docs/learn/journalhttps://posttrainllm.com/docs/learn/llm-mechanics-fundamentalshttps://posttrainllm.com/docs/learn/mac-mastery-maphttps://posttrainllm.com/docs/learn/model-vs-agenthttps://posttrainllm.com/docs/learn/qwen3-vl-mrope-deepstackhttps://posttrainllm.com/docs/learn/READMEhttps://posttrainllm.com/docs/learn/rl-multi-agent-roadmaphttps://posttrainllm.com/docs/learn/session-01-neural-net-basicshttps://posttrainllm.com/docs/learn/session-02-gradient-descenthttps://posttrainllm.com/docs/learn/session-03-non-linearitieshttps://posttrainllm.com/docs/learn/session-04-ml-paradigmshttps://posttrainllm.com/docs/learn/session-05-scalinghttps://posttrainllm.com/docs/learn/session-06-tokenization-embeddingshttps://posttrainllm.com/docs/learn/session-07-behavior-learninghttps://posttrainllm.com/docs/learn/session-08-training-mechanicshttps://posttrainllm.com/docs/learn/session-09-tensorshttps://posttrainllm.com/docs/learn/session-10-attentionhttps://posttrainllm.com/docs/learn/session-11-evals-rewardshttps://posttrainllm.com/docs/learn/small-model-tool-calling-playbookhttps://posttrainllm.com/docs/learn/speech-and-systems-topicshttps://posttrainllm.com/docs/learn/tool-calling-frontier-parityhttps://posttrainllm.com/docs/learn/webgpu-execution-modelhttps://posttrainllm.com/docs/learning-pipelinehttps://posttrainllm.com/docs/learning-progresshttps://posttrainllm.com/docs/learning/new-thingshttps://posttrainllm.com/docs/lm_eval_integrationhttps://posttrainllm.com/docs/lora_guidehttps://posttrainllm.com/docs/mac_app_planhttps://posttrainllm.com/docs/MAPhttps://posttrainllm.com/docs/memory_tradeoffshttps://posttrainllm.com/docs/model_guidehttps://posttrainllm.com/docs/moehttps://posttrainllm.com/docs/mtphttps://posttrainllm.com/docs/NEXThttps://posttrainllm.com/docs/noteshttps://posttrainllm.com/docs/online_softmax_in_attentionhttps://posttrainllm.com/docs/optimizershttps://posttrainllm.com/docs/pace-handoff-2026-06-10https://posttrainllm.com/docs/pace-landing-drafthttps://posttrainllm.com/docs/pace-model-manifesthttps://posttrainllm.com/docs/parked/ane-coremlhttps://posttrainllm.com/docs/parked/app-polishhttps://posttrainllm.com/docs/parked/browserhttps://posttrainllm.com/docs/parked/READMEhttps://posttrainllm.com/docs/parked/tier5https://posttrainllm.com/docs/parked/vlmhttps://posttrainllm.com/docs/peft_variantshttps://posttrainllm.com/docs/perf_audit_mlxfast_tiedhttps://posttrainllm.com/docs/perf_questhttps://posttrainllm.com/docs/perf_researchhttps://posttrainllm.com/docs/performancehttps://posttrainllm.com/docs/PLANhttps://posttrainllm.com/docs/planner-lock-2026-06-19https://posttrainllm.com/docs/prds/5.1-reasoning-on-22Mhttps://posttrainllm.com/docs/prds/5.2-testtime-compute-scalinghttps://posttrainllm.com/docs/prds/5.3-vision-language-toyhttps://posttrainllm.com/docs/prds/5.4-diffusion-lm-microhttps://posttrainllm.com/docs/prds/5.5-sparse-moe-kernelshttps://posttrainllm.com/docs/prds/5.6-tts-toyhttps://posttrainllm.com/docs/prds/5.7-explainer-video-modelhttps://posttrainllm.com/docs/prds/A1-first-specialist-tool-callerhttps://posttrainllm.com/docs/prds/B1-second-specialist-shell-or-sqlhttps://posttrainllm.com/docs/prds/B2-B7-router-familyhttps://posttrainllm.com/docs/prds/B5-cloud-escalate-traininghttps://posttrainllm.com/docs/prds/B6-mac-app-demohttps://posttrainllm.com/docs/prds/B8-multilingual-specialisthttps://posttrainllm.com/docs/prds/B9-energy-per-tokenhttps://posttrainllm.com/docs/prds/B10-quality-classifierhttps://posttrainllm.com/docs/prds/B11-wsd-schedulehttps://posttrainllm.com/docs/prds/B12-loss-spike-recoveryhttps://posttrainllm.com/docs/prds/B13-interp-on-checkpointshttps://posttrainllm.com/docs/prds/B14-speculative-decodinghttps://posttrainllm.com/docs/prds/B15-layerwise-lr-decay-sfthttps://posttrainllm.com/docs/prds/B16-m5-na-prefill-benchhttps://posttrainllm.com/docs/prds/B17-saelens-interophttps://posttrainllm.com/docs/prds/B18-nanochat-depth-knobhttps://posttrainllm.com/docs/prds/B19-group-saehttps://posttrainllm.com/docs/prds/B21-micro-automixerhttps://posttrainllm.com/docs/prds/B23-agent-eval-protocolhttps://posttrainllm.com/docs/prds/B25-scaledown-specialisthttps://posttrainllm.com/docs/prds/B26-deferred-toolshttps://posttrainllm.com/docs/prds/B28-composite-reward-frameworkhttps://posttrainllm.com/docs/prds/B31-gallery-and-project-pinshttps://posttrainllm.com/docs/prds/B32-eval-ci-gatehttps://posttrainllm.com/docs/prds/B33-laptop-finetune-onboardinghttps://posttrainllm.com/docs/prds/B34-batched-eval-runtimehttps://posttrainllm.com/docs/prds/B35-local-agent-vertical-pochttps://posttrainllm.com/docs/prds/C3-dora-ondisk-formathttps://posttrainllm.com/docs/prds/C4-tool-extractor-bpehttps://posttrainllm.com/docs/prds/C5-decode-jitter-thermalhttps://posttrainllm.com/docs/prds/C9-determinism-harnesshttps://posttrainllm.com/docs/prds/C10-train-run-dashboardhttps://posttrainllm.com/docs/prds/capability-retentionhttps://posttrainllm.com/docs/prds/continual-learning-loophttps://posttrainllm.com/docs/prds/E6-eval-scaledownhttps://posttrainllm.com/docs/prds/factory-completeness-trackerhttps://posttrainllm.com/docs/prds/factory-planner-v7-tools-in-prompthttps://posttrainllm.com/docs/prds/factory-vision-m4-architecture-decisionhttps://posttrainllm.com/docs/prds/factory-vision-m4-impl-planhttps://posttrainllm.com/docs/prds/factory-vision-specialisthttps://posttrainllm.com/docs/prds/game-rl-environment-pochttps://posttrainllm.com/docs/prds/gepa-prompt-evolutionhttps://posttrainllm.com/docs/prds/GPU-RESEARCH-BACKLOGhttps://posttrainllm.com/docs/prds/local-model-arena-selfplayhttps://posttrainllm.com/docs/prds/macos26-int8-ane-handoff-porthttps://posttrainllm.com/docs/prds/multi-turn-agentic-evalhttps://posttrainllm.com/docs/prds/pace-planner-v11-ship-gatehttps://posttrainllm.com/docs/prds/pace-planner-v11-training-datahttps://posttrainllm.com/docs/prds/pace-task-loop-v1https://posttrainllm.com/docs/prds/PRIORITYhttps://posttrainllm.com/docs/prds/qlora-large-model-finetunehttps://posttrainllm.com/docs/prds/quantized-inference-swifthttps://posttrainllm.com/docs/prds/READMEhttps://posttrainllm.com/docs/prds/self-improving-agentshttps://posttrainllm.com/docs/prds/specialist-pace-plannerhttps://posttrainllm.com/docs/prds/STATUShttps://posttrainllm.com/docs/prds/tinygpt-product-thesishttps://posttrainllm.com/docs/prds/vlm-ab-uivenus-vs-qwen3vlhttps://posttrainllm.com/docs/precisionhttps://posttrainllm.com/docs/progresshttps://posttrainllm.com/docs/PROJECT_RECOMMENDATION_CONTEXThttps://posttrainllm.com/docs/pruninghttps://posttrainllm.com/docs/qa_loghttps://posttrainllm.com/docs/quality_bundlehttps://posttrainllm.com/docs/quantization_expansionhttps://posttrainllm.com/docs/quickstarthttps://posttrainllm.com/docs/READMEhttps://posttrainllm.com/docs/recipes/automixhttps://posttrainllm.com/docs/recipes/b25-scaledownhttps://posttrainllm.com/docs/recipes/balanced-training-mixhttps://posttrainllm.com/docs/recipes/cookbook-character-specialisthttps://posttrainllm.com/docs/recipes/cookbook-personal-code-specialisthttps://posttrainllm.com/docs/recipes/cookbook-pydantic-aihttps://posttrainllm.com/docs/recipes/cookbook-smolagentshttps://posttrainllm.com/docs/recipes/distill-specialisthttps://posttrainllm.com/docs/recipes/distillation-fchttps://posttrainllm.com/docs/recipes/eval_plannerhttps://posttrainllm.com/docs/recipes/eval-gatehttps://posttrainllm.com/docs/recipes/from-traceshttps://posttrainllm.com/docs/recipes/mlx-exporthttps://posttrainllm.com/docs/recipes/pace-plannerhttps://posttrainllm.com/docs/recipes/READMEhttps://posttrainllm.com/docs/research/cross-stream-attention-evaluationhttps://posttrainllm.com/docs/research/indic_evalshttps://posttrainllm.com/docs/research/inference_benchmarks_may_2026https://posttrainllm.com/docs/research/mac_decode_baseline_m5prohttps://posttrainllm.com/docs/research/mac_slm_leaderboard_v0https://posttrainllm.com/docs/research/quality_benchmarks_may_2026https://posttrainllm.com/docs/research/wave_2_5_kernel_audithttps://posttrainllm.com/docs/research/wave_4_landscapehttps://posttrainllm.com/docs/RETROSPECTIVEhttps://posttrainllm.com/docs/roadmaphttps://posttrainllm.com/docs/roadmap/blockershttps://posttrainllm.com/docs/roadmap/categorieshttps://posttrainllm.com/docs/roadmap/datasetshttps://posttrainllm.com/docs/roadmap/honest_summaryhttps://posttrainllm.com/docs/roadmap/north_star_refinedhttps://posttrainllm.com/docs/roadmap/phased_planhttps://posttrainllm.com/docs/roadmap/recommended_orderhttps://posttrainllm.com/docs/roadmap/tier1https://posttrainllm.com/docs/roadmap/tier2https://posttrainllm.com/docs/roadmap/tier3https://posttrainllm.com/docs/roadmap/tier4_skiphttps://posttrainllm.com/docs/roadmap/tier5_frontier_2026https://posttrainllm.com/docs/session_2026_05_31https://posttrainllm.com/docs/session_retrospectivehttps://posttrainllm.com/docs/sessions/2026-06-05-eval-firsthttps://posttrainllm.com/docs/sessions/2026-06-06-mac-specialist-platformhttps://posttrainllm.com/docs/sessions/2026-06-06-specialist-embedder-blockerhttps://posttrainllm.com/docs/sessions/2026-06-06-vllm-mlx-investigationhttps://posttrainllm.com/docs/sessions/2026-06-08-factory-validation-nighthttps://posttrainllm.com/docs/sessions/2026-06-13-market-landscape-mac-firsthttps://posttrainllm.com/docs/sessions/2026-06-16-distill-to-self-improvementhttps://posttrainllm.com/docs/sessions/2026-06-17-stepback-inventory-roihttps://posttrainllm.com/docs/sessions/2026-06-17-vercel-eve-and-local-agentshttps://posttrainllm.com/docs/specialist_v1_findingshttps://posttrainllm.com/docs/specialists/a1-tool-callerhttps://posttrainllm.com/docs/specialists/b1-sqlhttps://posttrainllm.com/docs/specialists/b1-sql-pochttps://posttrainllm.com/docs/speculative_headshttps://posttrainllm.com/docs/streaming_llm_kivihttps://posttrainllm.com/docs/study_guidehttps://posttrainllm.com/docs/techniques/audit-inventoryhttps://posttrainllm.com/docs/techniques/method-vs-recipehttps://posttrainllm.com/docs/techniques/READMEhttps://posttrainllm.com/docs/techniques/sql-technique-backloghttps://posttrainllm.com/docs/techniques/trainloop-teardownhttps://posttrainllm.com/docs/test-coveragehttps://posttrainllm.com/docs/tool_call_extractorhttps://posttrainllm.com/docs/traininghttps://posttrainllm.com/docs/training_guidehttps://posttrainllm.com/docs/training/dpohttps://posttrainllm.com/docs/training/pretrainhttps://posttrainllm.com/docs/training/sfthttps://posttrainllm.com/docs/v11-baselines-2026-06-09https://posttrainllm.com/docs/validation_reporthttps://posttrainllm.com/docs/wwdc-2026-impacthttps://posttrainllm.com/docs/yoco_resultshttps://posttrainllm.com/eval-leaderboardhttps://posttrainllm.com/inferencehttps://posttrainllm.com/leaderboardhttps://posttrainllm.com/playgroundhttps://posttrainllm.com/recipeshttps://posttrainllm.com/roadmaphttps://posttrainllm.com/sae-timelinehttps://posttrainllm.com/speeduphttps://posttrainllm.com/training-dashboardhttps://posttrainllm.com/webgpu-test