7f59912916
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.
16 lines
781 B
TOML
16 lines
781 B
TOML
[package]
|
|
name = "makepad-ai-common"
|
|
version = "0.1.0"
|
|
edition = "2021"
|
|
description = "Cross-family internals shared by the AI model family crates: DiffusionError, sharded safetensors, the gpu_* / try_* exec surface, progress hooks, host .pth / Metal-GEMM helpers."
|
|
license = "MIT"
|
|
|
|
[dependencies]
|
|
makepad-ai-loader = { path = "../../loader" }
|
|
# Must not be target-specific: when this crate is a path dep of another
|
|
# standalone `[workspace]` (diffusion / family crates), cargo 1.92 drops
|
|
# target-cfg deps. metal-precompile bakes the metallib at build time so
|
|
# the first Metal dispatch is not a 13k-line shader compile.
|
|
makepad-ai-cuda = { path = "../../cuda" }
|
|
makepad-ai-metal = { path = "../../metal", features = ["metal-precompile"] }
|
|
makepad-ai-llm = { path = "../../llm" }
|