> Canonical page: https://posttrainllm.com/recipes

the recipe log

# Recipes, not vibes.

A recipe names the failure mode, the data, the eval gate, the slice gate, and the stop rule. "Try DPO" is not a recipe — it's a wish. These are the named, reproducible post-training and packaging recipes the factory runs, plus cookbooks for real integrations. Companion logs: the [devlog](https://posttrainllm.com/devlog) and the[experiment lab](https://posttrainllm.com/experiments).

## Post-training recipes

[start here Method vs recipe — the standard Why "try DPO" isn't a plan. A recipe names the failure mode, the data, the eval gate, the slice gate, and the stop rule. Open recipe →](https://posttrainllm.com/docs/techniques/method-vs-recipe/)[Distill a specialist from a bigger model Teacher → student on a local machine. KL + NLL mix, temperature, α, rejection sampling on a checker. Open recipe →](https://posttrainllm.com/docs/recipes/distill-specialist/)[Distill function-calling: Phi-3-mini → tiny The full function-calling distillation arc — the recipe behind the frontier-parity result. Open recipe →](https://posttrainllm.com/docs/recipes/distillation-fc/)[From .atraj rollouts to a trained specialist Turn agent trajectories into SFT data: filter, dedupe (MinHash), emit ChatML, train. The self-improving loop. Open recipe →](https://posttrainllm.com/docs/recipes/from-traces/)[eval-gate as a CI / pre-commit gate Wire a frozen eval into CI so a regression exits non-zero. The discipline that keeps main honest. Open recipe →](https://posttrainllm.com/docs/recipes/eval-gate/)[Balanced training mix Mixing corpora without one domain drowning the others — ratio search with a micro-AutoMixer. Open recipe →](https://posttrainllm.com/docs/recipes/balanced-training-mix/)[Micro-AutoMixer for data-mix ratios Search data-mix ratios cheaply before committing compute to a full run. Open recipe →](https://posttrainllm.com/docs/recipes/automix/)[ScaleDown — extractive context compression A specialist that compresses context extractively — smaller prompts, same answer. Open recipe →](https://posttrainllm.com/docs/recipes/b25-scaledown/)[packaging MLX export Package a trained adapter/model for the MLX runtime — the last mile before it serves. Open recipe →](https://posttrainllm.com/docs/recipes/mlx-export/)

## Cookbooks — real integrations

[Character specialist Train a small model with a consistent voice/persona, end to end. Open cookbook →](https://posttrainllm.com/docs/recipes/cookbook-character-specialist/)[Personal code specialist A code model tuned on your own repos and conventions. Open cookbook →](https://posttrainllm.com/docs/recipes/cookbook-personal-code-specialist/)[posttrainllm + Pydantic AI Drive a local posttrainllm model from Pydantic AI via the OpenAI-compatible endpoint. Open cookbook →](https://posttrainllm.com/docs/recipes/cookbook-pydantic-ai/)[posttrainllm + smolagents Run smolagents against a local specialist — agents on one machine, no cloud. Open cookbook →](https://posttrainllm.com/docs/recipes/cookbook-smolagents/)

## Complete technique → recipe registry

Every retained technique has a final project disposition and a concrete lab contract: target, failure mode, data, method, eval, regressions, budget, stop rule, decision rule, exercise, and mastery gate. These are ready learning labs, not an unfinished AI backlog.

[validated with caveat Constrained generation A capable model emits invalid syntax or forbidden output shapes. Lab Run a tiny legal-token or JSON-schema fixture and compare raw versus constrained validity. Mastery Explain why legal output is not the same as correct output. Open canonical recipe →](https://posttrainllm.com/recipes/constrained-generation)[validated with caveat Specialist distillation A small base lacks a narrow behavior that a qualified teacher can demonstrate. Lab Trace the file-ops lineage from stock 4B through distilled, multibackend, and ReST candidates. Mastery State the target gain and breadth cost that force routing. Open canonical recipe →](https://posttrainllm.com/recipes/distillation)[reference only Evolution strategies The objective cannot provide useful gradients or the model/policy is treated as a black box. Lab Simulate one two-parameter ES update on a deterministic toy reward. Mastery Explain the evaluation-cost tradeoff versus backpropagation. Open canonical recipe →](https://posttrainllm.com/recipes/evolution-strategies)[validated with caveat GaLore and training stability Optimizer state or unstable gradients block an otherwise valid training recipe. Lab Compare the recorded GaLore smoke with the gradient-checkpointing result and identify what each actually saves. Mastery Distinguish optimizer-state memory from activation memory. Open canonical recipe →](https://posttrainllm.com/recipes/galore-stability)[reference only Interpretability probes Aggregate scores cannot explain which layers, tokens, or representations drive a failure. Lab Inspect one logit-lens or activation-patching fixture and write the causal claim it does and does not support. Mastery Separate correlation, localization, and causal intervention. Open canonical recipe →](https://posttrainllm.com/recipes/interpretability)[validated with caveat LoRA, DoRA, and adapter geometry The base misses a bounded task and full fine-tuning is unnecessary or too costly. Lab Run or inspect LoRA geometry for one successful and one failed adapter. Mastery Explain rank, target modules, effective rank, and why lower loss is insufficient. Open canonical recipe →](https://posttrainllm.com/recipes/lora)[validated with caveat Sparse mixture of experts A dense model lacks capacity and a true sparse execution path is available. Lab Calculate active versus total parameters for top-k expert routing and identify why dense execution saves nothing. Mastery Explain the difference between sparse parameters and sparse compute. Open canonical recipe →](https://posttrainllm.com/recipes/moe)[reference only Multi-token prediction Single-token supervision or decode throughput is the measured bottleneck. Lab Draw the label shift for two future-token heads and compute the added loss terms. Mastery Explain why training more heads does not guarantee faster decoding. Open canonical recipe →](https://posttrainllm.com/recipes/mtp)[validated with caveat Optimizer and schedule selection Loss spikes, under-updating, or optimizer memory blocks a frozen recipe. Lab Diagnose three recorded loss curves as learning-rate, capacity, or data failures. Mastery Choose an optimizer change from evidence rather than novelty. Open canonical recipe →](https://posttrainllm.com/recipes/optimizers)[validated with caveat Parameter-efficient fine-tuning variants Plain LoRA has a measured geometry, capacity, quantization, or many-adapter limitation. Lab Create a one-page comparison of LoRA, DoRA, QLoRA, and VeRA parameter/memory assumptions. Mastery Select a PEFT variant from the bottleneck, not from its name. Open canonical recipe →](https://posttrainllm.com/recipes/peft-variants)[validated with caveat Precision and numerical gates Precision choice causes overflow, drift, excess memory, or slow execution. Lab Measure fp32/fp16 error on a tiny matrix operation and choose a tolerance before looking at the result. Mastery Explain why a faster kernel with wrong numerics is not an optimization. Open canonical recipe →](https://posttrainllm.com/recipes/precision)[validated with caveat Unstructured and structured pruning Artifact size or dense topology is the measured deployment bottleneck. Lab Compare 50% zero weights with physical layer removal and predict which metrics change. Mastery Distinguish statistical sparsity, compressed storage, and runtime sparsity. Open canonical recipe →](https://posttrainllm.com/recipes/pruning)[validated with caveat Quantization and storage formats Full precision exceeds the target's distribution or memory budget. Lab Trace one fp16-to-int4 artifact from calibration through load and decode measurement. Mastery Explain why four-bit storage can still execute with wider arithmetic. Open canonical recipe →](https://posttrainllm.com/recipes/quantization)[validated with caveat Speculative heads and decoding Autoregressive decode latency is the measured user-visible bottleneck. Lab Compute expected speedup from draft length, acceptance rate, and verification cost. Mastery Identify when speculative decoding becomes slower than ordinary decoding. Open canonical recipe →](https://posttrainllm.com/recipes/speculative-heads)[validated with caveat StreamingLLM and KIVI cache compression KV cache growth, not model weights, is the measured memory blocker. Lab Calculate full, windowed, int8, and int4 KV bytes for one model/context. Mastery Choose between eviction and quantization from the actual context failure mode. Open canonical recipe →](https://posttrainllm.com/recipes/streaming-kivi)[closed lineage SQL specialist lineage Small-model SQL gains fail to generalize across public and synthetic domains or emit prose wrappers. Lab Reconstruct the SQL lineage and mark which variable changed at each step. Mastery Explain why routing beat blending and why DPO failed to remove the wrapper. Open canonical recipe →](https://posttrainllm.com/recipes/sql-lineage)[closed experiment Needle 45M successor factorial Tiny overfit can pass while held-out selection remains below stock and destructive confirmation behavior fails completely. Lab Reconstruct the three factorial effects, apply the logical early-stop rule, and design the next 1.7B capacity test without changing the public gate. Mastery Explain why 100% tiny overfit coexists with a conclusive failed promotion and why sealed V2 stayed closed. Open canonical recipe →](https://posttrainllm.com/recipes/needle2-evaluation)[closed experiment Parakeet WGSL browser ASR Browser audio ingestion may be too slow, too large, nondeterministic, or network-dependent. Lab Re-score the two prediction files, identify which WhisperKit errors come from numeric normalization, and compare aggregate with median per-clip RTFx. Mastery Explain why 0.0% WER and a 3.51x paired latency win can coexist with a correct reject decision under the frozen 50x gate. Open canonical recipe →](https://posttrainllm.com/recipes/parakeet-browser-asr)

## Technique registry

[SQL technique lineage The closed ledger of SQL recipes and their final dispositions — each tied to its reward surface. Open →](https://posttrainllm.com/docs/techniques/sql-technique-backlog/)[TrainLoop teardown What was learned (and stolen) from an external post-training platform. Open →](https://posttrainllm.com/docs/techniques/trainloop-teardown/)[Audit 2026 technique inventory Row-level treatment of every technique in the 2026 audit: keep, experimental, flagged, delete. Open →](https://posttrainllm.com/docs/techniques/audit-inventory/)

 Need the full training loop? [Read the MLX LoRA and QLoRA field guide →](https://posttrainllm.com/mlx-lora-fine-tuning)