6cdc923813
Some checks failed
wai cross-architecture byte-equality / windows-x86_64 (push) Successful in 31s
wai cross-architecture byte-equality / linux-aarch64 (push) Successful in 1m7s
joule-code conformance gate / Build + round-trip C1-C4 vectors (push) Failing after 1m24s
WAI conformance / cargo-class corpora · windows (push) Successful in 1m13s
WAI conformance / cross-arch byte-equality · windows (push) Successful in 22s
jouleclaw public surface must stay documented / Public-surface audit (jouleclaw-rs) (push) Successful in 1m4s
wai cross-architecture byte-equality / linux-x86_64 (push) Successful in 22s
WAI conformance / cargo-class corpora · linux-arm64 (push) Successful in 2m4s
WAI conformance / cargo-class corpora · linux (push) Successful in 36s
WAI conformance / cross-arch byte-equality · linux (push) Successful in 21s
WAI conformance / cross-arch byte-equality · linux-arm64 (push) Successful in 1m16s
WAI neural decoders end-to-end / synthetic fixtures → Rust decode → reference (±4) (push) Failing after 1m17s
wai cross-architecture byte-equality / macos-arm64 (reference) (push) Successful in 5m56s
WAI conformance / cargo-class corpora · macos (push) Successful in 12m7s
WAI conformance / cross-arch byte-equality · macos (push) Successful in 36s
Five workflows install an exact toolchain instead of deferring to rust-toolchain.toml (wai-conformance, wai-neural-e2e, wai-byte-equality, joule-code-conformance, jouleclaw-public-surface-documented). Align them with the now-uniform 1.98.0 dev/root pin so CI builds the same compiler as the dev box. Co-Authored-By: Claude Opus 4.8 <[email protected]>
63 lines
3 KiB
YAML
63 lines
3 KiB
YAML
name: WAI neural decoders end-to-end
|
|
|
|
# The float neural decoders (elic / mlicpp / glc / dcvc / snac) cannot make
|
|
# WAI's cross-architecture byte-equality promise — their ONNX synthesis is
|
|
# float, so the low bits differ by hardware (this is exactly why the
|
|
# integer-exact int_synth path exists, and why only it is in the byte-equality
|
|
# matrix). What they CAN be is *correct end-to-end* on a real machine: entropy
|
|
# decode → ONNX synthesis → image assembly, reproducing a reference within ORT
|
|
# float rounding (±4). This job proves that, with no committed model blobs:
|
|
# the committed tool builds minimal synthetic CompressAI-compatible models +
|
|
# bitstreams, and the Rust decoders reproduce each reference. It runs only when
|
|
# the neural decode code or the generator changes — not on every cargo class.
|
|
|
|
on:
|
|
push:
|
|
branches: [main]
|
|
paths:
|
|
- 'wai/wai-rs/src/neural/**'
|
|
- 'wai/wai-rs/tests/neural_native_decode.rs'
|
|
- 'wai/tools/wai_validate_synthetic.py'
|
|
- '.github/workflows/wai-neural-e2e.yml'
|
|
pull_request:
|
|
paths:
|
|
- 'wai/wai-rs/src/neural/**'
|
|
- 'wai/wai-rs/tests/neural_native_decode.rs'
|
|
- 'wai/tools/wai_validate_synthetic.py'
|
|
- '.github/workflows/wai-neural-e2e.yml'
|
|
|
|
jobs:
|
|
neural-e2e:
|
|
name: synthetic fixtures → Rust decode → reference (±4)
|
|
runs-on: linux
|
|
steps:
|
|
- uses: actions/checkout@v5
|
|
- uses: actions/setup-python@v5
|
|
with:
|
|
python-version: '3.11'
|
|
- name: ML stack (CPU torch + torchvision + compressai + onnxruntime)
|
|
run: |
|
|
python -m pip install --upgrade pip
|
|
# torch AND torchvision from the CPU index so their compiled ops match
|
|
# (compressai imports torchvision; a mismatched build fails at
|
|
# `torchvision::nms does not exist`).
|
|
pip install torch torchvision --index-url https://download.pytorch.org/whl/cpu
|
|
# `onnx` (the exporter) is separate from `onnxruntime` (the decoder).
|
|
pip install compressai onnx onnxruntime numpy pillow
|
|
- name: Build the synthetic fixtures (elic / mlicpp / snac / dcvc / glc)
|
|
working-directory: wai
|
|
run: python tools/wai_validate_synthetic.py --size 256
|
|
# Bootstrap rustup: the docker runner image ships no Rust.
|
|
- name: Install Rust 1.98.0
|
|
run: |
|
|
export PATH="$HOME/.cargo/bin:$PATH"
|
|
command -v rustup >/dev/null 2>&1 || \
|
|
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y --profile minimal --default-toolchain 1.98.0
|
|
rustup toolchain install 1.98.0 --profile minimal
|
|
rustup default 1.98.0
|
|
echo "$HOME/.cargo/bin" >> "$GITHUB_PATH"
|
|
- name: Each Rust decoder reproduces its reference end-to-end
|
|
working-directory: wai/wai-rs
|
|
# --no-default-features: the neural decoders need only `ort`, NOT the C
|
|
# codec chain (AV1/JXL/opus), so the job needs no system codec deps.
|
|
run: cargo test --no-default-features --features neural --test neural_native_decode -- _matches_reference --nocapture
|