Repository navigation
feat(benchmarks): add reproducible CIFAR-style Transformer-G / ResNet18-feature-D benchmark - #586
Merged
Merged
Conversation
…18-feature-D benchmark
…ir, register cifar_toy in Forge history
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Adds a fast (~1.5–4.5 min/run on 1× RTX A6000), reproducible CIFAR-style 32×32 image toy benchmark (
benchmarks/cifar_toy/) and contract tests (tests/test_cifar_toy_benchmark.py) matching the structural regime of HyperGAN'scifar10-tiny-transformer, using only ParticleGAN's public API (GANTrainer,get_recipe,MoGParticlePrior,particlegan.init.deterministic_orthogonal_).Benchmark Structure
TinyTransformerGenerator32): 2-block 64-token Transformer -> 32×32 RGB (16matrix/conv params,2,489,731trainable params).ResNet18FeatureCritic32): Frozen SHA-256-verified ImageNet ResNet-18 (layer1/2/3at 64px, strictly locked ineval()mode) + 3 trainable feature heads (9matrix/conv params,171,779trainable params). Zerotorchvisiondependency.ResNet34FeatureEvaluator): Independent frozen SHA-256-verified ImageNet ResNet-34 512-dimavgpoolFeature-FID, unbiased polynomial KID (×1e3), and 10-class centroid mode coverage/TV (real-vs-real CIFAR-10 oracle floor:FID=82.66,KID×1e3=61.2±206.3,10/10cov,TV=0.088).AGENTS.mdcompliant): Protocol seed0, deterministic public init, isolated critic/generator data + trainer RNG streams, bit-identicalinitial_g_sha256,initial_d_sha256,initial_prior_sha256, andbatch_sequence_digestacross trainers, zero task-host repairs, zero bulk logs in Git. Registered"cifar_toy"inexperiments/forge/history.pywith refreshed Forge catalog.Runnable CLI
Matched Baseline Leaderboard (Real CIFAR-10, 1× RTX A6000, Seed 0,
steps=800,B=128)Step-0 FID:
2977.78(N=200) /2965.82(N=16384). Real-vs-Real oracle FID:82.66.bispf_optimized_p200_b128_s800bispf_optimizedempf_p200_b128_s800empfempf_p16384_b128_s800empfbispf_optimized_p16384_b128_s800bispf_optimizedunified_v11_p16384_b128_s800unified_v11unified_v11_p200_b128_s800unified_v11Verification
python scripts/check_lint.py— PASS (294baseline, 0 additions)python -m pytest -q tests/test_cifar_toy_benchmark.py tests/test_forge_history.py— 19 passedpython -m experiments.forge validate && python -m experiments.forge compile --check— PASS (CURRENT,11861/11861)