Taken from vjroger/work. A full merge of that fork fights this work branch
in widgets; this is the stems crate, vocal mel-band, span-cache lanes,
log_ring, output fence, midi inject, effect_doc, mp3 sniff, mp4 audio
edit, and audio-only file open. Stage on origin/main cargo-checks again.
libs/sqlite_query reads WITHOUT ROWID tables through the index cursor,
refuses to write them, and the pager releases its process write slot on
drop. libs/tar is an in-repo tar reader with gzip through fast_inflate.
libs/git imports packed objects with bounded checkout writers and its
HTTP file responses carry a trailer. libs/code_language gains a Haskell lexer with literate (Bird) dialect
detection. libs/loader_bundle and libs/search are new;
libs/app_module carries the super-app module surface; libs/workspace
adds Settings.infinite_zoom and RendererChoice::gpu_env_value; libs/ai
builds without warnings across the hub, llm, metal and model crates;
windows-rs job object handles are c_void.
Squashed from work (the libs parts of each):
- Restore retained renderer support for Scope (libs/search)
- Share Builder target across Makepad app builds (libs/loader_bundle)
- Index local Apple Mail with Gmail labels, attachments and reimport (libs/sqlite_query)
- libs/ai: warning cleanup across the hub, llm, metal and model crates
- code_language: a Haskell lexer with literate (Bird) dialect detection
- git: packed imports and bounded checkout writers; HTTP file responses carry a trailer
- builder: dedicated-folder installer checks, catalog and runtime fixes; Windows job objects hold c_void handles (libs/windows)
- workspace: Settings.infinite_zoom, the experimental prepared map inside the glyph
- wm: the Android super-app hosts apps as on-demand dylibs (libs/app_module)
- libs/tar: an in-repo tar reader; the super-app unpacks its archives with it
- workspace: RendererChoice::gpu_env_value follows the platform's runtime GPU choice
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
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>
The 5090 kept about 1.5 GB of CUDA allocator pool after evicting every
model, so a card that fits flux2-dev refused it (30510 MB free reported,
32090 MB in a fresh process). Admission now trims the pool when the last
resident is gone and re-measures before refusing; usable VRAM is measured
after the same trim; the refusal names the pool it released.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The per-family model crates (flux, h3, paint, music, speech, stems,
vision, sfx, rife, trellis, beats, common) and the libs/diffusion
research binaries read some 190 environment variables that were
research knobs: tensor dumps, per-stage timing, oracle-parity and
fixture rigs, experiment toggles. The path taken with none of them set
is the one that ships; every such knob is deleted with the code it
gated, and every losing branch of an experiment toggle goes with its
toggle — dead kernels, fields and functions included. What remains are
the real configuration variables (the FLUX_*_MODE family, FLUX_GRAPH,
the VAE pool cap, the FLUX2 text-encoder residency, H3_VAE_BATCH, the
music3 caches and official modes, the stems/beats f16 switches, the
weight and data roots) and the build-script variables.
Rebased on the runtime cleanup: precision stays explicit everywhere
(GemmPrecision, f16_attention_operands, the H3 text precision, DA3's
StrictF32 in code); no act16, no H3_ACT_F16, no FLUX_ATTN_F16 or
FLUX_VAE_CONV_GEMM reads survive.
Reviewed by the delegate reviewer (APPLY, no findings) and gated on the
Windows CUDA box: all seventeen model crates check, motion and vision
tests, the hub and the diffusion bins — the gate caught one CUDA-only
tap marker the Mac never compiles, removed here. On this Mac: the same
checks plus motion 24, paint 158 and vision 23 tests.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The AI runtime crates read their precision, activation and kernel
choices from environment variables; the code path taken with none set
is the one that ships, so every knob that selected it is now an explicit
argument and every losing path is gone. GemmPrecision { f16_accumulate,
f16_activations } is passed by each caller: the default {true, true} is
the old unset Flux route; H3's DiT and text encoder pass {false, false}
(H3's >1e4 activations saturate f16 — the policy H3 used to set through
FLUX_GEMM_F16ACC=0 on itself), its VAE {true, false}; DA3's StrictF32
selects f32 packed attention in code; Hy-Motion carries an explicit
f16_attention_operands flag through its text refiner, its double and
single blocks and the CUDA backend (true in production, false only in
its full validator). The libs/diffusion bins — a separate workspace —
are migrated to the same shapes.
Benches and validators no longer set variables on themselves: llama's
skip-logits is a session option (the CUDA bench turns it on), OCR takes
explicit use_f16_gemm and tiled_roformer options, the lane speculative
probe reads its CLI. The live gates the first cut had deleted are back
as explicit-config tests: MMQ M=129, the strided-f32 MMV path, the
RMS+MUL CPU oracle. The loader's THREADS and CHUNK_MB stay real settings.
The Metal quantized-matmul experiment (metal_qmm and its vendored MLX
kernels) was reachable only through a knob and goes with it.
Reviewed in three rounds by the delegate reviewer (the last round
accepted everything but one Hy-Motion call site, fixed in round four
and reviewed here), and gated on the Windows CUDA box: lib checks of
common/paint/loader/cuda/llm/motion/vision, motion 24 and vision 23
tests, the hub check, the diffusion bins, llm 253 passed / 1 ignored.
On this Mac: the same checks plus the motion and vision tests.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
LocalModels runs registry models in-process (install state, resumable downloads, recorded licence acknowledgements at $MAKEPAD_HOME/license_acks.json, weight paths by file role) and libs/ai/hub_ui is the install panel + licence modal every app can embed. New native ports: Beat This! (beats + downbeats) and Basic Pitch (note transcription) with their registry entries; the Salamander Drumkit samples (CC BY-SA 3.0, 37 files pinned by size and sha256) as a sample bank the downloader fetches like a model.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The sam3dbody backend reads its options from the request's prompt string
(`hands`, `detect`, `persons=N`): `hands` runs the full mode and the
packet carries which hands were fused and their boxes; `detect` finds up
to N persons with SAM 3.1 (an optional native-segment role on the body
entry, the same artifact the segment entry pins) and runs one body pass
per person with its box and mask, so the packet's people array grows.
The body crate shares one body pass between the packet, mask and hands
entry points, and infer_full takes the mask prompt too.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
BodyModel::infer_masked takes a provided person box and its full-frame 0/1
mask: the mask is warped with the crop's own affine (bilinear, rounded back
to 0/1 like the reference's uint8 warp), encoded by the prompt encoder's
mask CNN (four stride-2 convolutions with channel LayerNorm and erf GELU,
then a 1x1 to 1280) and added to the backbone tokens before the ray
conditioning, replacing the folded no-mask term. Oracle parity on the
mask fixture: warp exact, every CNN stage within f32 accumulation noise,
conditioned context 8e-4, end to end kp3d 2.2 mm / kp2d 0.25 px. 41 tests.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
BodyModel::infer_full runs the reference's full mode natively: the body
pass's hand boxes become two 512 crops (the left one cut from the mirrored
image, padding 0.9), each goes through the backbone, its own ray
conditioning and the hand decoder (the body decoder with the *_hand
tensors), the hand-mode rig pre-transform (local_to_world_wrist, wrist and
root offsets, non-hand parameters zeroed) and the hand camera head's
scale factor of 10; the left result is un-mirrored. The fusion gates each
hand (wrist angle, box size, keypoint spread, wrist distance), re-prompts
the body decoder with the trusted wrists and elbows as point prompts (the
decoder now takes N prompt tokens and a previous estimate), and writes the
fused wrist angles, hand parameters and hand scale/shape back before the
final rig pass. pose.rs gains the roma xyz (extrinsic) and XZY (intrinsic)
euler pairs; mhr.rs returns joint global rotations.
Oracle parity on both full-mode fixtures: every hand-decoder stage per
step (tokens 7e-3, heads 4e-4, rig params 1e-4), fusion reports as the
reference (one hand trusted on the standing photo, both on the crop), and
end to end kp3d 1.6 mm / kp2d 0.85 px. 36 tests.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
One backend call per SAM-style decoder layer (PE norms, token
self-attention, token-to-image cross-attention, erf-GELU feed-forward,
final norm) inside one command buffer, with the layer's f32 weights cached
on the device under their content identity and pooled transients.
gpu_two_way_layer_resident in the common backend; CUDA declines and the
per-op path stays. Body decoder loop on the M3 Max: layers 100 -> 9 ms,
frame 405 -> 266 ms, oracle parity unchanged (kp3d 1.5 mm).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The Metal tensor backend is host-Vec based: every op copied its inputs up
and its result back, which put the DINOv3 backbone at 2.1 s a frame on an
M3 Max. try_vit_backbone_resident_f32 runs a whole pre-norm ViT stack
(LayerNorm, q/k/v, rotate-half rope from tables, flash attention, out,
LayerNorm, SwiGLU, down; residuals in place) inside one command buffer
against cached bf16 weights, with a new kernel_makepad_rope_half_tables_f32
kernel for the rope. The precompiled metallib now carries the bf16 GEMM
kernels the runtime source compile already enabled on bfloat devices.
The common backend exposes it as gpu_vit_backbone_resident (CUDA declines;
its per-op path is already resident) and the body backbone tries it first:
233 ms a frame on the M3 Max, same oracle parity (kp3d 1.5 mm).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The FP8 backbone mode added each linear's bias by downloading, tiling and
re-uploading it per call, which cost more than the FP8 GEMM saved. A
gpu_add_cols_broadcast op (CUDA kernel; host loop on the Metal tensor
backend) adds a cols-wide bias to every row on the device, and each linear
keeps its bias resident after the first upload.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
BodyModel::set_backbone_fp8 quantises every backbone weight to E4M3 with
a per-tensor absmax scale and runs the tensor-core FP8 GEMM (bias
broadcast after). A backend without FP8 turns it off per layer on the
first refusal, so Metal keeps bf16. The oracle test reports its
accuracy and timing next to bf16; the default stays bf16 until the
numbers say otherwise.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The backbone's linears now take the cuBLASLt bias-epilogue / bf16 mm
paths (bf16 operands, f32 accumulation, bf16 output — the reference's
precision) with the f32-accumulating GEMM as the fallback. The rig's
pose correctives run on the final refinement step only by default: the
intermediate steps only feed keypoints back into the decoder, and the
oracle shows the difference is 0.4 mm (2.1 vs 1.7 mm against the
reference) for half the loop's rig time; `correctives_every_step` keeps
the exact mode.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The backbone, conditioning and decoder now take any square crop whose
side is a multiple of 16 (512 is the trained size); BodyModel::set_crop_size
selects it and caches the dense positional grid per size. Measured on the
oracle image against the reference: 512 gives 1.7 mm mean keypoint error,
384 and 256 about 2 cm mean (6 to 7 cm worst joint), 192 falls apart —
the knob is a real accuracy trade, not free speed. The decoder loop
reports its split (layer chain, heads, rig+camera, refinement) so the
next optimisation is chosen on numbers.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Apps ask the hub for a recognizer or a voice and get one; where it runs is
the hub's decision. AiHub::start_stt / start_tts return poll-driven
sessions shaped like the chat session. The Auto ladder is Whisper/Kokoro in
this process (weights present, machine election), on the machine node over
loopback, on a LAN node, else the OS engine; SpeechReach::Local is the
"don't reach out" knob. Audio always comes back as PCM: the app owns the
device.
Three layers:
- makepad-ai-speech is the whole speech model family, engines only.
libs/voice (Whisper + Silero VAD) folds in as the `whisper` and `vad`
modules next to kokoro and indextts, each a cargo feature; the Apple
bridges and the Speaker/VoiceTranscriber selection leave it.
- makepad-system-speech (new) is the OS speech services as blocking fns:
Apple SpeechAnalyzer/AVSpeechSynthesizer via Swift, Windows.Media.Speech*
on the vendored bindings, Android SpeechRecognizer/TextToSpeech through
MakepadSpeech.java (API 26 floor), espeak-ng on Linux. It models the two
STT shapes honestly: PCM in (Whisper, Apple) versus an engine that owns
the microphone (Android, Windows), with capabilities the caller reads.
- the hub grows speech sessions, in-process Whisper/Kokoro workers with the
residency election, a `whisper` wire backend (stt domain, registry entry
pinned to ggerganov/whisper.cpp) so a Mac can serve a Quest, and a
`language` field on the generate request.
Consumers: the Window voice input runs on an STT session and switches to
engine-mic mode when the recognizer owns the microphone; converse's
SpeechOutput is a lazily started TTS session plus a pump thread; route
drops its private speech copy for converse; vj's lyrics fallback and the
alignment bakes call the engines directly.
Verified here: speech-roundtrip through the real sessions (Apple voice in,
in-process Whisper on Metal out, 4.3% WER); system-speech-test TTS->STT
verbatim; hub/converse/system-speech unit tests; msvc, aarch64-android and
linux-gnu cross-checks; Java against android-34. Windows, Android and Linux
bridges are compile-checked only.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The backbone (20 heads) and the decoder (8 heads) both have 64-wide
heads, which the stack's FA2 kernel covers: f16 operands with f32
softmax and accumulation, the reference's own precision class, instead
of the composite path that materialises the 1029 x 1029 scores per head.
The composite path stays as the fallback where a backend lacks the
kernel. Oracle parity unchanged.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The ray-conditioning conv used to download the 1024 x 1280 embedding,
concatenate the 99 ray features on the host and upload the result; it is
now two resident linears over the two column blocks with the no-mask
term folded into the bias, plus an add (32 ms -> a few on the 4090).
The decoder loop downloaded the whole normalised token block every
layer for its one pose row; it now slices that row and fetches the
block once at the end (or per layer under a trace). Same numbers.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
On the 4090 the frame was 885 ms of which 840 were host scalar loops in
the decoder loop: the three 70-row refinement FFNs (about 240M
multiply-adds a step), the two heads, and the rig's 55k x 3000
corrective output layer plus its identity blendshape sum. Those are now
GPU-resident linears (heads.rs GpuStepHeads, MhrRig::prepare_gpu), the
skinning computes one transform per joint instead of one per influence,
and the results are identical (all 29 oracle tests unchanged). The rig
still runs entirely on the host when no GPU side was prepared.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
model.rs closes the loop: crop -> backbone -> ray-conditioned context ->
six decoder steps, each through the pose head, the rig, the camera and
the projection, then the packet the sandbox already reads (kp3d/kp2d in
camera axes, the 204 rig parameters, global rotation, camera translation,
joint positions). Against the reference on the oracle image, on Metal:
3D keypoints within 1.7 mm, 2D within 0.4 px, rig parameters, camera
and rotation within 2e-3. packet.rs writes the JSON by hand with the
reference worker's rounding and field order.
The hub gains the `body-native` feature (default on): registry entry
`sam3dbody` pinned to the Comfy-Org repack by revision, size and sha,
body_native_backend.rs beside the subprocess reference backend with the
same live_step contract, the `body` capability advertised when the
feature is compiled, and a stubbed test double for the CPU-only tests.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
decoder.rs builds the 145-row token set (pose, previous, prompt, two
hand-box rows, 70 keypoint rows, 70 3D-keypoint rows), runs the six
layers with the SAM-style repeated positional encoding (none on the
first layer's self-attention), cross-attends against the ray-conditioned
context, and after each step hands the normalised pose token to the
heads and updates the keypoint rows from the caller's feedback: the
2D-keypoint positional FFN and the bilinearly sampled context features
on the valid rows, the pelvis-centred 3D positional FFN on the rest.
heads.rs holds the host-side ReLU FFN heads, the refinement FFNs, the
hand-box MLP and the hand classifier. Against the oracle on Metal, every
layer's residual stream is within 6e-3, every step's pose head within
8e-4 and camera head within 2e-5; token assembly and the host heads
match to 1e-4.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
dino.rs is the 32-block ViT-H+/16 on the gpu_* surface after the TRELLIS
conditioner: bf16-resident linears with f32 accumulation, layer scale
folded into the output projections, rotate-half rope, SwiGLU, the
model's own final norm. preprocess.rs is the crop (box -> 1.25 pad ->
3:4 -> square -> 512 bilinear warp, ImageNet normalisation), the CLIFF
condition vector and the patch rays; condition.rs is the dense
positional encoding and the ray-conditioned decoder context. Against the
reference oracle on Metal: backbone 1.1% mean relative (bf16 noise),
ray-conditioned context 8e-5, crop within one u8 rounding step, rays and
dense PE 1e-7. One reference detail the paper does not state: its shrink
of the ray field is an antialiased filter whose taps clip at the image
edge, so the two edge patches sample inside their block centre (9.03 and
501.97 rather than 7.5 and 503.5); block centres left 0.1 of error.
The Metal tensor backend was missing rope_half (it aliased the
interleaved layout) and layer_norm_mul_add; both now exist with the
CUDA contract, which is what lets this backbone run on Apple silicon.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
mhr.rs is the Momentum Human Rig on the CPU: blendshapes, the 889x249
parameter transform, parents-first similarity kinematics (x,y,z,w
quaternions, Rz Ry Rx eulers, 2^s scales), the sparse-then-dense pose
corrective MLP over joints 2.., linear blend skinning and the 308-row
keypoint regression. pose.rs decodes the 519-wide head output: 6d
rotations, the 23 ball / 58 hinge / 6 translation body layout, the
mixed-dof hand layout, scale and hand component bases, the camera
translation and the perspective projection. Against the reference
oracle: vertices within 1e-4 cm with correctives on, keypoints within
1e-6 m, rig parameters within 1e-7. Two conventions the spec could not
settle on paper are now settled by the fixture: the head's global
rotation triple arrives Z,Y,X-ordered, and the corrective features start
at joint 2 (750 wide). Fixture tests skip cleanly without the oracle
directory or the weights.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
makepad-ai-body joins the AI model workspace: the constants of the
architecture (DINOv3 ViT-H+/16 at 512, the 6-layer promptable decoder,
the 519-wide pose head, the 127-joint / 18439-vertex MHR rig) and the
single-file safetensors reader for the Comfy-Org repack, which fails
closed on a Meta checkpoint-style header and checks the shapes the port
is written for at load. The backbone, decoder and rig modules follow
in their own lanes against the spec under local/agent_state/sam3dbody.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Squashed from work:
- asset-ai: FastH3 4-step fast video backend; clip keyframes on the wire
- asset-ui: loop video chains — text→image→video that ends where it began
- h3: safetensors -> pruned-Q4_K GGUF quantizer for the 24GB DiT tiers
- h3_quant_gguf verify: row-error gates calibrated to the measured Q4_K floor
- asset-ai realtime: the feedback loop — the source anchors, the drifted frame inits
- asset-ai realtime: a feedback loop survives a resize and travels by default
- asset-ai realtime: the feedback loop frees itself from the feed handshake and pauses for its listener
- asset-ai realtime: the outbound encode leaves the loop's critical path
- asset-ai ocr: the ocr domain — Chandra 2 at page resolution, and the tower goes planner-owned
- llm slots: a lane can hold an image span — embedding prefill and a rope cursor of its own
- vision tower on CUDA: the encode leg gets its two missing kernels
- llm/ocr: one M-RoPE grid encoder for both image paths, and a livelock made an error
- vision tower on CUDA: the f16 GEMM keeps the precision it was throwing away
- live: a feed that moves box takes its trip with it — one seed image
- vision tower on CUDA: the tiled attention becomes bit-exact, and tensor cores go
- llm prefill on CUDA: the MMA attention kernel gets the tile a 4-to-1 model needs
- asset-ai ocr: the CUDA encode lane joins the integration — vision-parity sits beside run's three arms, and the kernels
- Merge branch 'ocr-perf-integration' into work
- asset-ai: the live anchor can follow the trip, and text leaves the 5090
- asset-ai: the camera moves the world, and the world starts still
- asset-import: the EA strategy classics, in the one 2D contract
- rtsmap: one seeded generator for tiled strategy maps
- asset-ui: one card for the strategy classics, with a pack dropdown
- asset-ai: music3 reference-audio path, ocr/h3 backends, registry
- asset: mp4 sample index for range-streaming, chat tools, import profiles
- cnc: tiberium is twelve growth frames, not twelve empty variants
- platform: native file and save dialogs, in-house on all three desktops
- chat: the scan holds out for a lane home
- chat: a full home queues you — take the free lane
- chat: the preload has a percentage, and the boundless cap stops showing
- llm cuda: the 32x2 attention tile — even GQA ratios stay on MMA
- sa3 gets a bake path: the sfx model's tables precomputed by a diffusion-side bin
- sqlite_query: anti-join regression test
- td import: HARV's second frame block is its harvesting cycle, not a turret
- asset-ui: sprite enhancement runs on the 32B dev DiT — distillation, not the prompt, was the ceiling
- ai-hub: makepad-asset-ai becomes makepad-ai-hub at libs/ai/hub, the chat pane becomes makepad-chat-ui, the service bin
- asset-ui: test health fixtures grow the realtime field they were born without
- ai-hub: one home at ~/.makepad — weights/ run/ cache/ logs/, the service cache migrates from ai_content by a single re
- ai-hub: subprocess workers die with the node — process groups everywhere, PDEATHSIG on linux, one KILL_ON_JOB_CLOSE Jo
- ai-hub: the hub object — AiHub::in_process, pipes vocabulary, and the local LLM engine generalized out of mpfiles (aic
- strict-json: the dependency-free JSON module gets its own crate; asset-client re-exports it so nothing downstream move
- ai-hub: the machine layer — node entries, the 0600 machine token, and the residency election that IS the lock (aicore
- ai-hub: MPHUB1 — the fabric beacon only dedicated nodes can send (aicore §4)
- ai-hub: job leases — work lives only while it is renewed (aicore §8)
- asset-creator: the pipeline library is born — specs, the deps gate, and the derived-state law (aicore §9)
- ai-hub: RAM residency facts — the CPU-side twin of residency.rs (aicore §3)
- ai-hub: ETA placement primitives — relative GPU throughput, the four-term estimate, and an observable breakdown (aicor
- ai-hub: leases go live on the wire — origin fields on submit, /job/<id>/keepalive, /bye, and the reaper that cancels w
- ai-hub: the chat providers move in — fleet qwen, openai, grok, claude/codex/grok CLIs, the responses driver, and the w
- asset-creator: the engine — one pipeline run against the hub, deps-gated, spliced, cancellable, resumable-by-construct
- ai-hub: the machine node mode — --machine binds loopback, registers in ~/.makepad/run, and exits on its own once idle
- asset-creator: makepad-creator-run — the detached client for runs that must outlive a window (aicore §9)
- ai-hub: a native Claude Messages-API provider — API-key or Claude Code OAuth, bounded SSE streaming, injected tools (a
- route + converse: off makepad_ai — the Agent seam moves to converse, route's cloud dispatcher rides the hub's Claude p
- asset-creator: the preset tables move in — fifteen chain-policy constants shared by every creator app (aicore §9 / P6)
- makepad_ai is deleted — every backend is a hub pipe, the agent seam lives with its consumers (aicore §14, decided 2026
- ai-hub: loads hold the machine residency election — set_model_state claims on Loaded and publishes the service port (a
- ai-hub: chats run the machine election — route to a serving holder, wait on a loading one, claim and publish when open
- ai-hub: pick_for_domain_eta — ETA-ranked placement over the shared hard-filter core (aicore §6 / P4)
- asset-creator: the engine picks a provider per stage at dispatch time — a chain's later stages see fresh fleet state (
- ai-hub: the fabric secret gates the service HTTP surface — bearer on everything but /health and the ticketed peer path
- vj: DREAM runs execute in the app — pipelines.rs becomes the run it used to watch (aicore §9 / F1)
- asset-creator: the runner — generate one thing and put it in the catalog, one implementation for every surface (aicore
- chat-ui: the session runs in the app — no broker anywhere on the chat path (aicore P8 / F5)
- asset-store: assets.query is a first-class query endpoint — the bounded SQL surface outlives the broker (aicore P8 / F
- asset-creator: CreatorTools — the chat tool pack for a store that only stores (aicore §9 / P8)
- asset-store: the shrink — the store stores (aicore P7)
- importer + asset-server host: the coordination era ends (aicore P7)
- store config purge + asset-ui goes fleet-direct; the derive protocol gets its route proof (aicore P7)
- client + chat dispatcher: the dead wire comes out (aicore P7/P8)
- ai-hub: 0.3.0 — the health version says which era a node runs
- ai-hub: the default fleet is 'gen' — apps hear the LAN without env plumbing
- ai-hub: the preload note percents the prefill, not the job bar
- ai-hub: conversations keep their KV — the wire mirror, the lane identity, the in-turn dynamic context (aicore §7)
- ai-hub: an open-think model is thinking from its first token
- libs: the zero-warning sweep — stitch casts say what they mean, xatlas keeps upstream's surface quietly
- zero-warning sweep, round two — the first full-workspace pass
- zero-warning sweep, round three — the model lanes and the deep examples
- zero-warning sweep, round four — the last stragglers
- zero-warning sweep, round five — vj and chat-ui
- zero-warning sweep, round six — three cascades
Co-authored-by: Claude <info@makepad.nl>
Squashed from work; the fine-grained history is under tag archive/work-2026-08-29:
- mp* wave: mpwm window manager + the mp app family, WM API, theme bridge, PDF engine fix
- work: land the sources the last commits reference
- kenney: catalogue all 50 free 3D kits; Modal dismissed() never fired
- asset store: central vision-annotation queue; Kenney donate prompt
- asset-ai: vision domain — image + prompt -> text on every fleet node
- annotation runs the fleet's vision services through the normal job queue
- vj: responsive DJ mixer + Windows drag-and-drop, cherry-picked from PR #1199 (vjroger)
- store search: WordNet synonym expansion, query-side; per-term seeks
- video scheduling: a cold model pin never downloads past a warm one, and a dedicated box only serves its role
- an expansion can never lose a run — and `expand: true` finally means something
- fleet panel: a slow box is not a missing box
- h3: first+last keyframe conditioning (the weights were always FL2VA)
- registry: the three H3 FL2VA tiers name their real conditioning
- fleet scheduling: spread before stacking, and a stuck job moves
- a download never steals a job from a box that has the weights
- the faster GPU takes the tie: 6000 > 5090 > 4090
- an evicted flux model gives the card its VRAM back
- a job now records what each of its stages was handed
- open a stage in RUNS and read what it sent
- every expanded song was an instrumental: the lyrics had nowhere to go
- the writer never overwrites words the person wrote
- store: a dependent job's body is spliced from its deps' results at claim
- a fleet box's job row says what that job was asked for
- the store runs a whole pipeline, and one record says how far it got
- a client can declare a whole run, watch it, and stop it
- 100% now means published, and the expander is a job you can queue
- one card says what a spawned task is doing, everywhere
- a run that ends says so, instead of being noticed later
- when every box holding the model is busy, buy another copy
Squashed from work; the fine-grained history is under tag archive/work-2026-08-26:
- vj: the tweener learns — RIFE runs on the Mac and feeds the warp
- asset-ai: the chat tells the truth while it works
- llm: a cold turn on the solo slot takes the session-native speculative path — think-mode turns no longer re-ingest the whole conversation through the draft head (66 → 122 tok/s on the four-lane box)
- llm: the step cost model is chosen per device — the RTX PRO 6000's measured verify curve (13.7 + 3.17·B ms) beside the 5090's; the bench warms every tail shape and times two windows
- vj: local store, lyrics and model plumbing, and the frame-interpolator's device parity check
The model code was spread across eight crates that had grown into each other:
ggml and cuda and mlx each owned part of a tensor runtime, llama and tts and
voice2 each owned part of a model, and libs/diffusion owned everything else.
They are now one tree with an explicit shape:
libs/ai/cuda — kernels and launch surface
libs/ai/metal — Metal shaders and the shim
libs/ai/llm — the language-model runtime (sessions, lanes, contexts,
the CUDA and Metal executors, the compiled Metal path)
libs/ai/models/ — common, flux, h3, music, paint, speech, stems, vision
libs/diffusion is not deleted but demoted: what remains is the VALIDATOR
crate — several dozen `*_validate.rs` oracles that check a native
implementation against a reference, which is where they belong now that the
implementations live next door.
The functional work inside the move is mostly in the LLM runtime: N lanes that
draft while one verify batch serves all of them, per-slot prefill over a shared
folded attention arena, speculation that survives batching, and a scheduler
that reports rather than publishes. And in the CUDA build: a machine without
usable CUDA must still LINK (and say so), the default kernel arch is the
building machine's GPU, `NO_CUDA` forces the stub even where the toolkit
exists, and kernels compile in parallel with progress.
libs/video_flow is new here: classical optical flow estimation and the `mkfl`
motion-field payload — a flow field measured from a clip without a model,
which is what drives free-rate bounce-looping playback and the uprez/tween
enhance pipe.
