Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit History

Commits on Apr 13, 2026

• Update names: quantum-predecoder --> Ising-Decoding

  Show description for 615ce6b
  bmhowe23

  committed
  615ce6b

  Copy full SHA for 615ce6b

Commits on Apr 10, 2026

• fix(test): increase LER sample count to 100k to fix stochastic flakes on A100 (#60)

  Show description for a61c149
  

  ivanbasov

  and

  claude

  authored
  a61c149

  Copy full SHA for a61c149

Commits on Apr 8, 2026

• Add standalone script to generate evaluation test data (#57)

  Show description for 0925137
  wsttiger

  authored
  0925137

  Copy full SHA for 0925137

• fix: suppress double measurement noise injection in MemoryCircuit (#53)

  Show description for ed7b89a
  

  3 peopleauthored
  ed7b89a

  Copy full SHA for ed7b89a

• fix: find_best_model accepts named .pt files without epoch numbers (#55)

  Show description for 2a4e0f4
  

  ivanbasov

  and

  claude

  authored
  2a4e0f4

  Copy full SHA for 2a4e0f4

• Fix export of fp8 ONNX files (#52)

  Show description for 2cbb707
  

  3 peopleauthored
  2cbb707

  Copy full SHA for 2cbb707

Commits on Apr 7, 2026

• feat: rename pretrained models to Ising-Decoder-SurfaceCode-1-{Fast,Accurate} (#51)

  Show description for 812b935
  

  ivanbasov

  and

  claude

  authored
  812b935

  Copy full SHA for 812b935

• Perf imp for cudaq decoders (#42)

  Show description for ab688e0
  sacpis

  authored
  ab688e0

  Copy full SHA for ab688e0

• fix(mid): seed BitMatrixSampler explicitly to restore test reproducibility (#43)

  Show description for d09beb7
  

  ivanbasov

  and

  claude

  authored
  d09beb7

  Copy full SHA for d09beb7

Commits on Apr 3, 2026

• fix(ler): force num_workers=0 when torch.compile is active to prevent segfault (#41)

  Show description for 1be4aee
  

  ivanbasov

  and

  claude

  authored
  1be4aee

  Copy full SHA for 1be4aee

• feat(decoder_ablation): TRT pre-decoder backend + cudaq-qec docs and tests (#34)

  Show description for 8a884c6
  

  ivanbasov

  and

  claude

  authored
  8a884c6

  Copy full SHA for 8a884c6

• ci: add cu12/cu13 matrix to GPU unit tests (#36)

  Show description for 3ce44fe
  

  ivanbasov

  and

  claude

  authored
  3ce44fe

  Copy full SHA for 3ce44fe

Commits on Apr 2, 2026

• fix: guard against double torch.compile in SDR→LER path (#39)

  Show description for f6c7864
  kvmto

  authored
  f6c7864

  Copy full SHA for f6c7864

• fix: steps_per_epoch should count forward passes, not optimizer steps (#38)

  Show description for 31c60dd
  

  huaweil-nv

  and

  ivanbasov

  authored
  31c60dd

  Copy full SHA for 31c60dd

• feat(ci): add multi-GPU tests and CI job for DDP validation (#37)

  Show description for 2c6f22e
  

  ivanbasov

  and

  claude

  authored
  2c6f22e

  Copy full SHA for 2c6f22e

Commits on Apr 1, 2026

• Improvement cuStabilizer DEM sampling backend and consolidate tests (#32)

  Show description for 5aeebdf
  

  kvmto

  and

  bmhowe23

  authored
  5aeebdf

  Copy full SHA for 5aeebdf

Commits on Mar 25, 2026

• ci: exercise multi-worker DataLoader (num_workers > 0) in CI (#29)

  Show description for 1b49b69
  kvmto

  authored
  1b49b69

  Copy full SHA for 1b49b69

• fix: remove sdr_as_percent kwarg from compute_syndrome_density() call (#30)

  Show description for 12895ce
  

  ivanbasov

  and

  claude

  authored
  12895ce

  Copy full SHA for 12895ce

Commits on Mar 24, 2026

• Add cuStabilizer BitMatrixSampler integration to DEM sampling (#24)

  Show description for a250ca1
  

  3 peopleauthored
  a250ca1

  Copy full SHA for a250ca1

Commits on Mar 23, 2026

• Replace proprietary license headers with Apache-2.0 (#19)

  Show description for b5ff9eb
  

  ivanbasov

  and

  claude

  authored
  b5ff9eb

  Copy full SHA for b5ff9eb

• Remove unused model architecture (#18)

  Show description for af9da6e
  bmhowe23

  authored
  af9da6e

  Copy full SHA for af9da6e

• Add global decoder ablation workflow (#17)

  Show description for 9f6d854
  sacpis

  authored
  9f6d854

  Copy full SHA for 9f6d854

• V3 optimizations from development branch (#13)

  Show description for 2e99b0e
  

  kvmto

  and

  ivanbasov

  authored
  2e99b0e

  Copy full SHA for 2e99b0e

Commits on Mar 19, 2026

• Add ONNX export, INT8/FP8 quantization, and SafeTensors inference loading (#14)

  Show description for 993e797
  

  3 peopleauthored
  993e797

  Copy full SHA for 993e797

Commits on Mar 11, 2026

• Add training scripts and configs for public model generation (#4)

  Show description for 65e12c0
  

  ivanbasov

  and

  claude

  authored
  65e12c0

  Copy full SHA for 65e12c0

Commits on Mar 9, 2026

• Add fp16/fp32 SafeTensors export and inference loading support (#11)

  Show description for e70d9ab
  

  ivanbasov

  and

  IgorBaratta

  authored
  e70d9ab

  Copy full SHA for e70d9ab

• feat(qec): surface-code noise model upscaling for training (6e-3 target) (#2)

  Show description for 1c62bc7
  ivanbasov

  authored
  1c62bc7

  Copy full SHA for 1c62bc7

Commits on Mar 7, 2026

• Add CI test tiers: mid-running and long-running GPU tests (#3)

  Show description for 90804e0
  ivanbasov

  authored
  90804e0

  Copy full SHA for 90804e0

Commits on Mar 6, 2026

• Add YAPF lint config, reformat codebase, and enforce in CI (#10)

  Show description for 037ef8e
  ivanbasov

  authored
  037ef8e

  Copy full SHA for 037ef8e

• Remove legacy PreDecoderModelMemory_v1.0.94.pt, migrate to r9/r13 models (#5)

  Show description for da3ab52
  ivanbasov

  authored
  da3ab52

  Copy full SHA for da3ab52

Commits on Mar 5, 2026

• fix(test): stabilize flaky LER tests in CI (#6)

  Show description for cd578cc
  ivanbasov

  authored
  cd578cc

  Copy full SHA for cd578cc

Commits on Mar 4, 2026

• Widen d=9 LER improvement tolerance to 5e-4

  Show description for 62cf726
  ivanbasov

  committed
  62cf726

  Copy full SHA for 62cf726

• Widen d=13 LER range checks to prevent CI flakes

  Show description for b84321f
  ivanbasov

  committed
  b84321f

  Copy full SHA for b84321f

• Widen LER improvement tolerance from 5e-5 to 1e-4

  Show description for 20e9a92
  ivanbasov

  committed
  20e9a92

  Copy full SHA for 20e9a92

• Initial commit using cf02b459 from private repo

  Show description for 36cae25
  ivanbasov

  committed
  36cae25

  Copy full SHA for 36cae25

Loading