makepad/libs/ai/models/notes/tests/transcription.rs
Admin 6ba1b9425c libs: ai hub/models/cuda, speech, chat_ui
Squash of 54 work commits (Sep 1–12):
  6251f7c  ai-hub: body domain — live pose packets ride the realtime session
  ea50c77  chat_ui: the feed's session gets its profile brief back
  f51b5f3  ai-body: the crate for the native SAM 3D Body port, with its weights reader
  8211ae6  ai-body: the MHR rig and the pose head's parameter decoding, oracle-exact
  9e343a8  ai-body: the DINOv3 ViT-H+/16 backbone, crop and ray conditioning; Metal gains rope-half and affine layer norm
  69d842c  ai-body: the promptable pose decoder and its refinement loop, oracle-matched on Metal
  66e5e2f  ai-hub: SAM 3D Body runs natively — `sam3dbody` on the body domain, oracle-matched end to end
  a634198  ai-hub: the body-native commit carried a peer's in-flight hub hunks; put them back where they were
  9ff44e8  ai-hub: the body-native wiring, this time only the lane's hunks
  6a1c16b  ai-body: third-party notices — what the port is implemented after, and what it is not
  d78411a  ai-body: the per-step work moves to the GPU
  b22259b  ai-body: the context stays on the GPU; only the pose token leaves the loop
  346f31f  ai-body: flash attention for the head-dim-64 blocks
  45b5b98  ai-body: the crop size is a runtime knob, and the loop reports where its time goes
  4be6d19  ai-body: the test modules import the grid constants they still use
  7598346  ai-body: tensor-core GEMMs for the backbone, and the rig's correctives only where they count
  a9ce596  ai-body: the crop warp runs across cores
  8964ba6  ai-body: an FP8 backbone mode, off by default, measured against the oracle
  a2aaa8f  ai-body: the FP8 bias rides a column-broadcast add on the device
  d53c77d  metal: a device-resident ViT stack, and the body backbone rides it
  d006d0a  metal: resident f32 linears keep their weight on the device
  525ba1c  metal: a device-resident two-way decoder layer, and the body decoder rides it
  c9e6d88  ai-body: the hands pass — hand crops, the hand decoder, the hand-mode rig and the wrist fusion
  62dff26  ai-body: the mask prompt — a person's segmentation mask conditions the body pass
  a648cf8  ai-hub: body session options — hands, detect, persons=N
  8c568df  ai-hub: drop the SAM 3D Body reference worker backend
  7ff875a  ai-hub: keep a peer's in-flight beats/notes/local work out of the body commits
  31e5faa  ai-hub: local model runner, licence acknowledgements, a shared install panel; Beat This!, Basic Pitch and the Salamander drum-kit entries
  b94bc58  ai-services: the wire, the app port and the panel state — one conversation, many apps
  2acb798  ai-services: wire v2 — endpoints, receiver-side caps, result disposition
  8ae0ffb  ai-services: the engine core — registry, router and conversation, tested against a scripted model
  2308736  ai-services: the real models behind the engine feature — local through the hub, Claude, and none
  c3f631d  livepipe: one reusable pipe from a camera to a fleet node and back
  ff62db3  ai libs: the runtime env-var cleanup — precision is a per-caller policy, not an environment side channel
  04a94ef  realtime: one service-log line when a live session opens and one when it closes
  0ecb81c  ai models: the model-crates env-var cleanup — 172 research knobs gone, the unset default is the code
  4ca36c1  ai hub + services: the assistant's model comes from wherever it is resident — the fleet chat box, with tools, then the local weights
  432121e  aichat engine + wm: launch, then use — the assistant continues in the same turn once the app it started is on the bus
  7a5bf69  ai-hub registry: the Salamander drumkit samples come from the makepad.nl mirror — the GitHub repo only carries the .sfz files
  102ffc5  ai-services: messages on the bus — a manifest declares topics, the engine subscribes on a tool's behalf or by ToolResult.subscribe, a service publishes Message frames, an idle conversation wakes on a message as an event turn under rate laws; the WM bus forwards the new frames; every app that matches the wire gets its arm
  a837792  hub + flow: a whitespace-only chat completion is retried once and then fails instead of passing as an answer; a flow's model is a fleet model id unless it names a weight file on disk; chat models show under the text domain in /v1/models
  bc6c620  hub + flow: what the chat review found — the in-process route retries an empty completion too, a node says whether its prefill opened thinking so a brief-mode answer is never discarded, a preferred model falls back to normal election when no node has it, discovery keeps looking for the preferred model until patience runs out
  75c3441  hub: the PRO 6000 serves image as well as chat and text
  ad5e98b  hub registry: flux2-dev's VRAM estimate is its measured peak, 30 GB
  c7241e0  hub: a node that evicted every resident releases its cached allocator pool before refusing a load or publishing usable VRAM
  30575f0  flow: route generation by request workload
  1be1e21  ai-hub: gate downloads by disk capacity and recover fleet admission
  df6b394  filesystem_watcher, bounded_http, ai services: live and tool prerequisites
  79ebdb9  ai-hub: add a native Pixal3D image-to-3D backend
  0ba0d74  ai-hub: propagate typed refusals under reject queue policy
  cc6c872  Speed up H3 conditioning and video decoding
  e512059  Fix Qwen vision residency and generated material colors
  2864f68  ai-hub http client: bound every plain TCP connect to 3 s per address
  3d93229  ai: CUDA is a Linux/Windows-only dependency; the hub library defaults to llm + stt

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
2026-09-15 13:40:31 +02:00

121 lines
4.3 KiB
Rust

use makepad_ai_notes::config::{AUDIO_N_SAMPLES, SAMPLE_RATE};
use makepad_ai_notes::{NotesModel, MODEL_FILE};
use std::path::{Path, PathBuf};
fn checkpoint() -> PathBuf {
Path::new(env!("CARGO_MANIFEST_DIR"))
.join("../../../../local/models/weights/basic_pitch/nmp.onnx")
}
fn frequency(midi: f64) -> f64 {
440.0 * 2.0f64.powf((midi - 69.0) / 12.0)
}
fn attacked_sine(midi: f64, seconds: f64) -> Vec<f32> {
let samples = (seconds * SAMPLE_RATE as f64).round() as usize;
(0..samples)
.map(|index| {
let time = index as f64 / SAMPLE_RATE as f64;
let attack = (time / 0.008).min(1.0);
let release = ((seconds - time) / 0.015).clamp(0.0, 1.0);
(0.75 * attack * release
* (std::f64::consts::TAU * frequency(midi) * time).sin()) as f32
})
.collect()
}
fn has_note_near(notes: &[makepad_ai_notes::NoteEvent], midi: u8, start: f64) -> bool {
notes
.iter()
.any(|note| note.midi == midi && (note.start_secs - start).abs() <= 0.040)
}
#[test]
fn silence_produces_no_notes() {
let mut model = NotesModel::load(checkpoint()).unwrap();
let result = model.transcribe(&vec![0.0; AUDIO_N_SAMPLES]).unwrap();
assert!(result.notes.is_empty());
}
#[test]
fn four_note_bass_line_has_expected_pitches_and_onsets() {
let mut audio = Vec::new();
for midi in [28.0, 33.0, 38.0, 43.0] {
audio.extend(attacked_sine(midi, 0.5));
}
let mut model = NotesModel::load(checkpoint()).unwrap();
let result = model.transcribe(&audio).unwrap();
assert_eq!(result.notes.len(), 4, "unexpected bass notes: {:?}", result.notes);
for (midi, start) in [(28, 0.0), (33, 0.5), (38, 1.0), (43, 1.5)] {
assert!(
has_note_near(&result.notes, midi, start),
"missing MIDI {midi} near {start:.2}s; got {:?}",
result.notes
);
}
}
#[test]
fn c_major_triad_is_simultaneous() {
let voices = [attacked_sine(60.0, 0.7), attacked_sine(64.0, 0.7), attacked_sine(67.0, 0.7)];
let mut audio = vec![0.0f32; voices[0].len()];
for voice in voices {
for (sample, value) in audio.iter_mut().zip(voice) {
*sample += value / 3.0;
}
}
let mut model = NotesModel::load(checkpoint()).unwrap();
let result = model.transcribe(&audio).unwrap();
assert_eq!(result.notes.len(), 3, "unexpected triad notes: {:?}", result.notes);
for midi in [60, 64, 67] {
assert!(
has_note_near(&result.notes, midi, 0.0),
"missing triad MIDI {midi}; got {:?}",
result.notes
);
}
}
#[test]
fn one_semitone_glide_has_rising_bends() {
let seconds = 0.8;
let samples = (seconds * SAMPLE_RATE as f64) as usize;
let mut phase = 0.0f64;
let mut audio = Vec::with_capacity(samples);
for index in 0..samples {
let time = index as f64 / SAMPLE_RATE as f64;
let midi = 45.0 + time / seconds;
phase += std::f64::consts::TAU * frequency(midi) / SAMPLE_RATE as f64;
let envelope = (time / 0.008).min(1.0) * ((seconds - time) / 0.015).clamp(0.0, 1.0);
audio.push((0.75 * envelope * phase.sin()) as f32);
}
let mut model = NotesModel::load(checkpoint()).unwrap();
let result = model.transcribe(&audio).unwrap();
let note = result
.notes
.iter()
.filter(|note| note.midi == 45 || note.midi == 46)
.max_by(|a, b| a.end_secs.total_cmp(&b.end_secs))
.unwrap_or_else(|| panic!("missing gliding A; got {:?}", result.notes));
let reversals = note.bends.windows(2).filter(|pair| pair[1] < pair[0]).count();
assert!(reversals <= 1, "non-rising bend trend: {:?}", note.bends);
assert!(note.bends.last().unwrap_or(&0.0) > note.bends.first().unwrap_or(&0.0));
}
#[test]
fn overlap_seam_does_not_duplicate_a_sustained_note() {
let audio = attacked_sine(45.0, 2.4);
let mut model = NotesModel::load(checkpoint()).unwrap();
let result = model.transcribe(&audio).unwrap();
let long_a_notes = result
.notes
.iter()
.filter(|note| note.midi == 45 && note.end_secs - note.start_secs > 0.4)
.count();
assert_eq!(long_a_notes, 1, "seam duplicated A2: {:?}", result.notes);
}
#[test]
fn model_file_constant_matches_registry_cache_name() {
assert_eq!(MODEL_FILE, "basic_pitch_nmp.onnx");
}