[2/3] Add request sampling controls to all pinned Drama engines

CHANGELOG.md

@@ -4,6 +4,9 @@
 
 ### Fixed
 
+- CUDA NVFP4 long-form splitting preserves paired delivery tags, bracketed cues and special tokens,
+  including topic-shift handling, and recognizes CJK sentence/clause punctuation. An indivisible
+  protected span may exceed the character target; existing prompt and frame limits still apply.
 - Metal reference-prefill reuse now invalidates before incompatible/default prefill, compares the
   complete invariant prefix plus tower lookahead, preserves original chunk boundaries, and reuses
   one GPU-owned snapshot buffer pair per lane instead of allocating on every voice switch.
@@ -19,6 +22,13 @@
 
 ### Added
 
+- Per-request HTTP speech `temperature`, `top_p`, `top_k` and exact unsigned 64-bit
+  `seed` on all pinned profiles, with strict validation and propagation through
+  streaming, references and redraws. Omitted/null values retain existing defaults.
+- A checksummed CUDA NVFP4 engine patch series without changing its pinned upstream or weights.
+  Model-free build tests exercise real patch application, selection on Linux/macOS, corruption
+  refusal, repeat builds and preservation of modified checkouts. Manifests must cover each
+  selected patch exactly once; extra, missing and duplicate entries fail before checkout.
 - Native Apple-silicon support for the pinned NVFP4-pruned checkpoint: packed experts, compact
   original-ID routing and reference-conditioned semantic CFG, with independent backbone states.
   The Mac engine uses `engine/nvfp4/macos/UPSTREAM` and its checksum-verified patches; the existing

README.md

@@ -26,7 +26,7 @@
 
 [TensorFold](https://github.com/ashhart/TensorFold) runs Drama 3 natively since v1.1.0. This repository pins
 upstream TensorFold and adds a small series of opt-in patches for the full model. NVFP4 retains its
-measured CUDA engine pin and uses a separate upstream pin plus Metal patches on Mac. It builds the engine
+measured CUDA engine pin with a checksummed patch series and uses a separate upstream pin plus Metal patches on Mac. It builds the engine
 and starts a local speech server with an OpenAI-style `/v1/audio/speech` route, whole or streamed.
 
 **Built with Fish Audio.** The model weights are not in this repository; the launcher downloads them from
@@ -64,7 +64,8 @@ For the full model, write `--weights full` in both commands.
 The first start downloads the weights at the revision pinned in [`recipe/sources.json`](recipe/sources.json),
 into `drama-3-NVFP4-Pruned/` or `drama-3/`. Every start checks them against [`recipe/checksums`](recipe/checksums)
 and refuses missing or changed files; for the 21 GiB NVFP4 weights that takes about 11 seconds on a GB10.
-On Linux, `--weights nvfp4` builds [`engine/nvfp4/UPSTREAM`](engine/nvfp4/UPSTREAM) and serves with
+On Linux, `--weights nvfp4` builds [`engine/nvfp4/UPSTREAM`](engine/nvfp4/UPSTREAM) plus
+[`engine/nvfp4/patches`](engine/nvfp4/patches) and serves with
 guidance γ 1.5 and batched guidance, as measured; a `--cfg-gamma` of your own after `--` replaces both.
 On Mac it builds [`engine/nvfp4/macos/UPSTREAM`](engine/nvfp4/macos/UPSTREAM) plus its patch series.
 Metal's `--cfg-batched` pairs independent states in one command buffer, not a vectorized GEMM.
@@ -122,7 +123,9 @@ the license's Section III(c), which needs a separate license from Fish Audio.
 
 `POST /v1/audio/speech` takes `model` (`fishaudio/drama-3`), `input` (at most 4096 characters), `voice`
 (`default`, or a name loaded with `--voices`), `response_format` (`wav` or `pcm`) and `stream` (`true` needs
-`pcm`). Request-supplied references and other synthesis modifiers are refused. `GET /v1/voices` lists the voices.
+`pcm`). Optional `temperature`, `top_p`, `top_k` and exact integer `seed` select
+request-local sampling; see [ranges, defaults and redraw behavior](recipe/README-sampling.md).
+Other synthesis modifiers are refused. `GET /v1/voices` lists the voices.
 The Linux CUDA NVFP4 engine accepts its encoder's `drama3-voice` catalogs and request `references`
 as codec codes (not audio). The Mac NVFP4 engine instead uses attested `drama3-http-voice/1`
 catalogs and refuses request-supplied references; see the [Mac voice commands](recipe/README-macos.md#attested-reference-voices).
@@ -139,11 +142,31 @@ Guidance-enabled replies carry the configured γ in `X-Cfg-Gamma`; builtin voice
 
 ## The patches
 
-The patches build the full-model engine (`--weights full`). Upstream TensorFold already serves Drama 3. Each patch
+### CUDA NVFP4 long speech
+
+The CUDA NVFP4 patch series preserves delivery markup when `--long-form` splits a request.
+Paired tags such as `<whisper>…</whisper>`, bracketed cues and special tokens stay whole;
+Chinese/Japanese sentence and clause punctuation are recognized. A topic-shift mark inside a
+protected span does not create a cut. Plain English without these spans retains its existing cuts.
+
+`--long-form-chars` remains a code-point target, with one exception: an indivisible protected span
+may exceed it. The engine's prompt, segment-count and generated-frame limits still apply.
+The full-model and Metal NVFP4 engines do not expose this legacy long-form route.
+
+The behavioral reference is [Sogni-AI/drama3-mlx's chunking](https://github.com/Sogni-AI/drama3-mlx/blob/ecc235039ccd5f982f66a6982184f57432354ab4/drama3_mlx/chunking.py).
+The native patch extends TensorFold's own implementation; it does not copy that Fish-licensed source.
+See the [CUDA patch index](engine/nvfp4/patches/README.md) for scope and tests.
+
+### Full-model engine
+
+These patches build the full-model engine (`--weights full`). Upstream TensorFold already serves Drama 3. Each patch
 below adds one feature that stays off until you pass its flag, so a server started without them answers as
-upstream does. The exceptions are the first two: batched prefill is on by default (`TF_DRAMA_PREFILL_ROWS=1`
-restores upstream's single-row prefill) and the replay fix is a bug fix. The patches stack and apply in order;
-each one's notes are in the engine tree's `CHANGELOG.md` or `docs/`.
+upstream does. The exceptions are the first two and one request-level feature: batched prefill is on by default
+(`TF_DRAMA_PREFILL_ROWS=1` restores upstream's single-row prefill), the replay fix is a bug fix, and per-request
+`temperature`, `top_p`, `top_k` and `seed` (0013/0014) need no launch flag. A request without those fields is
+answered as upstream answers it; upstream refuses them with a 400, and these patches serve them. The CUDA NVFP4
+series (0002) and the Metal NVFP4 series (0004, 0005) carry the same request-level feature. The patches stack and
+apply in order; each one's notes are in the engine tree's `CHANGELOG.md` or `docs/`.
 
 | Patch | Adds | Opt in with | Upstream |
 |---|---|---|---|
@@ -159,6 +182,8 @@ each one's notes are in the engine tree's `CHANGELOG.md` or `docs/`.
 | 0010 (K6) | Whole replies that stopped short of their text are drawn again | `--short-reply-ratio`, `--short-reply-chars-per-s` | [#1161](https://github.com/ashhart/TensorFold/pull/1161) |
 | 0011 (K8) | A whole reply is drawn once more when a phrase opens far above the others | `--register-redraw-st`, `--register-redraw-high`, `--register-high-hz` | [#1162](https://github.com/ashhart/TensorFold/pull/1162) |
 | 0012 | One bounded reference-prefix snapshot per Metal lane, with exact identity and safe invalidation | none (automatic) | [#1237](https://github.com/ashhart/TensorFold/pull/1237) |
+| 0013 | Request-local speech sampling and exact integer seeds | HTTP `temperature`, `top_p`, `top_k`, `seed` | [#1277](https://github.com/ashhart/TensorFold/pull/1277) |
+| 0014 | Parsed-JSON floating boundary documentation and regressions | none (tests/docs) | [#1277](https://github.com/ashhart/TensorFold/pull/1277) |
 
 `drama3-voice`, a tool that writes attested voice files for `--voices`, is proposed upstream in
 [#1153](https://github.com/ashhart/TensorFold/pull/1153); it is not part of this series.

engine/nvfp4/UPSTREAM

@@ -1,5 +1,5 @@
-# The engine `scripts/build-engine.sh --weights nvfp4` builds, unpatched: the only engine that loads the NVFP4
-# pruned weights (plotarmordev/drama-3-NVFP4-Pruned). Commit b1f4b9b4 on branch drama-nvfp4 of the TensorFold fork;
+# The CUDA engine base for `scripts/build-engine.sh --weights nvfp4`, followed by patches/.
+# Commit b1f4b9b4 on branch drama-nvfp4 of the TensorFold fork loads the NVFP4 pruned weights;
 # results/full-vs-quant-01 measured the same source (v1.0.11-193-g70dec469; the two differ in comments and test
 # wording only).
 TENSORFOLD_REPOSITORY=https://github.com/plotarmordev/TensorFold.git