81c1813822
int_synth's determinism was proven on synthetic/constructed models; this adds a model TRAINED ON REAL IMAGES decoding a REAL PHOTOGRAPH, and proves the same bit-exactness on it. tools/wai_int_synth_real.py trains an autoencoder on real photos, post-training-quantizes the decoder with per-layer activation-scale calibration (the working recipe — pure QAT through the >> shift can't learn), encodes a held-out image to an int8 latent, and decodes it through the EXPORTED integer model in numpy (the same i64 scatter-accumulate / +bias / arithmetic >> shift / clamp that int_synth.rs runs) as the reference. New corpus vector `photo`: a 128×128 macOS Sonoma landscape → a 6,156-byte int8 latent (96×8×8) + a four-layer trained integer decoder, ~21.6 dB at ~3 bpp. Honest scope: a pure integer transposed-conv decoder with no entropy model — quality is modest BY DESIGN; the point is that the decode is bit-identical on every machine. wai_int_synth_vectors `gen` cross-checks the Rust decode against the python integer-forward (ref_rgb.bin) to max|Δ|=0 and freezes the rgb_hash; the four-platform conformance CI then enforces the same bytes everywhere. A real trained model, a real photo, decoded identically on every machine — the int_synth thesis, demonstrated end to end. Co-Authored-By: Claude Opus 4.8 <[email protected]>
9 lines
No EOL
131 B
JSON
9 lines
No EOL
131 B
JSON
{
|
|
"width": 128,
|
|
"height": 128,
|
|
"in_channels": 96,
|
|
"s_log2": 3,
|
|
"payload_bytes": 6156,
|
|
"psnr_db": 21.6,
|
|
"bpp": 3.006
|
|
} |