nisse/docs/Tuning-Guide.md

# Tuning Guide

How to fit a local model onto the inference host, and how to set the agent's
tuning knobs. Part 1 covers the ideas you need to reason about the hardware.
Part 2 covers each knob in `nisse-dev.yaml`. Part 3 gives an order to tune
in.

The knobs live under `tuning:` in the config file (see
`nisse-dev.example.yaml`). They are read at startup, so a change takes effect
on the next restart. Unknown or misspelt keys are startup errors.

---

## Part 1 — How a local model spends VRAM

A model server holding a model in GPU memory spends it on three things:

1. **Weights**, which are fixed for a given model and quantization.
2. **The KV cache**, which grows with context length.
3. **Working memory** for activations and runtime buffers. Budget 1–2 GB of
   headroom.

On a 24 GB card, the whole game is trading (1) against (2).

### Weights and quantization

A model's weights take roughly *parameters × bits per weight ÷ 8* bytes.

| | 16-bit (full) | 8-bit | ~4.5-bit (Q4_K_M, AWQ) |
|---|---|---|---|
| 24B | 48 GB | 24 GB | ~14 GB |
| 32B | 64 GB | 32 GB | ~19 GB |

Quantization stores each weight in fewer bits. Common 4-bit formats come out
nearer 4.5–5 bits per weight once scaling factors are counted, so check the
actual file size. The quality cost is small at 5–6 bits, usually acceptable at
4, and it climbs quickly below that. For agent work the cost shows up first as
**worse tool-call discipline**, not as worse prose. The capability matrix
(Part 3) measures exactly that.

**Mixture-of-experts** models such as Qwen3-30B-A3B are different. Memory is
set by *total* parameters (30B), but speed is set by the parameters *active*
for each token (3B). An MoE model of a given size therefore generates much
faster than a dense model of the same size. It's often weaker than a dense
model of the same total size, but not always for tool use.

### The KV cache, and why context costs memory

As the model reads the prompt, every layer stores a key and a value vector for
every token, so that later tokens can attend to them without recomputing.
That store is the KV cache. Its size per token is:

    2 (key and value) × layers × KV heads × head size × bytes per value

The numbers come from the model's `config.json` (`num_hidden_layers`,
`num_key_value_heads`, `head_dim`). Some worked examples at 16-bit:

| Model | Layers × KV heads × head size | Per token | 32k tokens |
|---|---|---|---|
| Qwen3-32B | 64 × 8 × 128 | 256 KB | ~8 GB |
| Mistral/Devstral Small 24B | 40 × 8 × 128 | 160 KB | ~5 GB |
| Qwen3-30B-A3B (MoE) | 48 × 4 × 128 | 96 KB | ~3 GB |

(Taken from the published configs; recheck against the model card before
relying on them.)

Two consequences:

- **Context length is a memory decision, not a free setting.** The spec's
  estimate (Part H) was about 5 GB left for KV cache beside a 4-bit 32B model.
  At 16-bit, that's only about 20k tokens of context for Qwen3-32B.
- **KV-cache quantization** stores keys and values in 8 bits (or 4). At 8
  bits it halves the cache and is close to lossless. At 4 bits it is
  noticeably worse. Quantizing the cache to 8-bit is usually the first lever
  to pull, since it buys twice the context for almost nothing.

The server allocates the cache for the context length you configure. It
doesn't matter how much of it a request uses. So set the context length on the
server deliberately: it is the ceiling for *prompt plus response* on every
request.

### Prompt processing vs generation

A request has two phases, and they run at very different speeds:

- **Prefill** reads the prompt. It is parallel and fast: thousands of tokens
  per second on a 4090.
- **Decode** writes the response one token at a time. Roughly 30–40 tokens per
  second for a dense 32B at 4-bit, and several times that for a 3B-active MoE.
  Treat these numbers as ballpark; measure your own.

An agent loop re-sends the whole transcript on every turn, which makes the
prompt long and the responses short (a tool call is often under 200 tokens).
So prefill matters more than the headline tokens-per-second suggests.

**Prefix caching** is the fix. If the start of a prompt matches the previous
request, the server reuses the KV cache for it and prefills only the new part.
llama.cpp does this per slot, and vLLM does it automatically. For this
workload it is often the single largest speed win. It works here because the
rendered transcript only grows at the end, so consecutive turns share
everything but their newest events.

### Tokens and characters

For English and code, a token is roughly **4 characters**. That's useful for
converting the character-based knobs below into context cost.

---

## Part 2 — The knobs

For each knob: what it controls, what goes wrong at either extreme, and what
to look at when deciding.

### `llm.max_output_tokens` — default 16384

The ceiling on a single response. It is sent with every request, for every
role.

- **Why it exists:** a response that hits the ceiling is truncated
  mid-sentence, or worse, mid-tool-call, where it can still parse. The agent
  detects truncation and treats it as an error, but a ceiling that's too
  tight still costs a failed turn. Planning with a frontier model was observed
  using over 4,000 output tokens.
- **Too high, on a local model:** the server reserves room for the whole
  response inside the context window. With a 32k context and a 16k ceiling,
  only 16k is left for the prompt. Some servers reject any request where
  prompt plus ceiling exceeds the context.
- **Too low:** truncated plans and tool calls.
- **Deciding:** executor turns are short, so 2–4k is plenty for a local
  executor. Frontier planning wants the large default. This one knob covers
  every role for now; a per-model setting is coming with the context budget
  (see *Coming next*).
- Environment override: `NISSE_MAX_OUTPUT_TOKENS`.

### `tuning.executor.max_turns` — default 24

How many model calls one milestone may make before the executor stops with
`capped` and asks you what to do.

- **Too low:** milestones that were making progress stop short. You'll see
  `capped` outcomes on work that was going well.
- **Too high:** a confused model keeps going. Every turn also adds its tool
  calls and results to the transcript, so a long run spends context as well
  as time.
- **Deciding:** look at how many turns *successful* milestones take, and set
  the cap comfortably above the high end, around 1.5–2× the turns of your
  longer successful runs. A stronger model usually needs fewer turns, not
  more.

### `tuning.executor.failure_limit` — default 2

How many consecutive failed `run_command` calls trigger escalation: the
sub-task is handed to a frontier model as a delegation (spec A12). A
successful command resets the count. Failed file reads don't count, since
reading is looking around, not doing the work.

- **Why it's a count and not a judgment:** small models are poorly calibrated
  about their own competence. Asked "are you stuck?", they say no until far
  too late, or yes constantly. A count of observable failures is a signal
  that doesn't depend on the model's opinion of itself.
- **Too low:** normal iteration escalates. Test-first work *expects* a failing
  test run before the fix, so a limit of 1 hands off routine work and spends
  frontier tokens on it.
- **Too high:** a weak model circles, re-running the same failing command
  with small variations, while the transcript fills up.
- **Deciding:** raise it as local models get stronger. Watch two things
  together. First, how often delegations fix what the local model couldn't
  (escalation is doing its job). Second, how often the local model recovers
  on its own after its second failure (the limit may be too low).

### `tuning.expert.consult_budget` — default 3

How many `ask_expert` calls one milestone may make. `ask_expert` is a single
stateless question to a stronger model, with no access to the repository
(spec A14), for things like "is this API deprecated?".

- **Too low:** the executor can't get unstuck on a knowledge question it
  can't answer from the code.
- **Too high:** a confused executor asks variations of the same question
  instead of doing the work or failing. That's the circling A12 warns about.
- **Deciding:** count consults per milestone, and check whether the answers
  changed what the executor did next. If it regularly exhausts the budget
  without progress, the budget is fine and the model or the plan is the
  problem.

### `tuning.delegation` — hard caps on one delegation

A delegation is a frontier model working a sub-task with full tool access
(spec A13). When a cap is hit, the delegation **fails**, and its branch and
work are discarded, because resuming on a half-edited tree isn't safe.

| Knob | Default | Caps |
|---|---|---|
| `max_turns` | 30 | model calls |
| `max_tokens` | 200000 | input + output tokens, as reported by the provider |
| `wall_clock_s` | 900 | elapsed time |

- **`max_tokens` is a scope limit, not the money limit.** Spending limits live
  on the LiteLLM key (spec A5). This cap stops one delegation from growing
  without bound. An unbounded one could still exhaust a daily budget and take
  every frontier model away until it resets.
- **Too low:** delegations fail just short of finishing and their work is
  thrown away. That is the most expensive way to be wrong, because you paid
  for the work.
- **Too high:** a stuck delegation burns tokens and time before failing.
- **Deciding:** every delegation records the tokens it used and how long it
  ran (the `delegation_completed` event, captured for evaluation per spec
  A16). Set each cap above what successful delegations actually use, with
  room to spare.

### `tuning.tools.command_timeout_s` — default 120

How long one `run_command` may run before its whole process group is killed.

- **Too low:** slow test suites, dependency installs and Docker builds get
  killed and reported as failures. That also counts toward `failure_limit`.
- **Too high:** a command that never exits blocks the turn for the whole
  timeout, such as a dev server started in the foreground or a test waiting on
  input. `stop` waits for a running command to finish too, so this is also the
  worst-case delay before a stop takes effect.
- **Deciding:** set it just above the slowest legitimate command in the
  projects the agent works on.

### `tuning.tools.max_output_chars` — default 4000

How much of a tool's output (command output, file contents, directory
listings) the model sees. Anything past this is cut, with a note saying how
much was dropped.

- **This is a context knob.** Every tool result stays in the transcript and is
  re-sent every turn. 4,000 characters is about 1,000 tokens, so 20 turns of
  maximal output is about 20k tokens, which is already most of a local
  model's context.
- **Too low:** the model doesn't see the part of the output that matters.
- **Too high:** the transcript outgrows the context window sooner.
- **A known weakness:** truncation keeps the *start* of the output, but test
  runners put their failure summary at the *end*. A long, failing pytest run
  can lose exactly the lines the model needed. Keeping the head and the tail
  would fix this, and is a candidate for the context-budget work.

---

## Part 3 — An order to tune in

Upstream choices change what's right downstream, so tune in this order and
change one thing at a time.

1. **Model and quantization.** Start with whether a candidate holds the
   tool-call contract at all:

       NISSE_CAPABILITY_MODELS=alias-a,alias-b \
         uv run pytest -m integration tests/integration/test_model_capability.py

   A model that falls back to prose instead of calling tools is disqualified,
   however good its benchmarks look. Before comparing aliases, check what
   each one actually points at in LiteLLM (`/model/info`, the
   `litellm_params.model` field). Two aliases believed to be the same model
   once weren't, and that cost a lot of debugging.
2. **Context length and KV-cache precision** on the server. Take the largest
   context that fits beside the weights with 1–2 GB of headroom. Try an 8-bit
   KV cache before giving up context.
3. **`llm.max_output_tokens`**, so the prompt keeps most of that context.
4. **Real milestones.** Run representative work and look at the evidence:
   turns used, escalations, delegation outcomes, and time taken.
5. **The loop knobs**: `failure_limit`, `max_turns`, `consult_budget`, the
   delegation caps, then the tool limits, one at a time.

**Expect noise.** Local model behaviour varies a lot from run to run. One
early comparison measured 12 of 20 runs holding the tool contract against 20
of 20, a real difference (p ≈ 0.003). Three runs each could easily have
pointed the other way. Before changing a knob because of a result, repeat the
run enough times that the result would survive a coin flip.

---

## Coming next

The next changes in the 4090-readiness work add knobs, and this guide gains a
section for each:

- **A context budget.** Rendering a transcript that fits the model's context
  window, with what to drop or summarize when it doesn't, and per-model output
  ceilings.
- **Enforced tool calls.** An explicit `finish` tool, so the server can be told
  a tool call is required and hold the model to the contract, where today it
  can fall back to prose.
- **A benchmark script** that runs the capability matrix and representative
  milestones across candidate models and prints a comparison table.