Search results

16 of 28 gradient-based optimizers lack IFusedOptimizerSpec, silently disabling fused compiled training

enhancementNew feature or requestNew feature or request

Status: Open.

#1930 In ooples/AiDotNet;

· ooples opened on Jul 30, 2026

Missing model: AWGIM (Attentive Weights Generation for Few-Shot Learning via Information Maximization)

enhancementNew feature or requestNew feature or request

Status: Open.

#1929 In ooples/AiDotNet;

· ooples opened on Jul 30, 2026

698 models train at the optimizer's default learning rate, not their paper's

Status: Open.

#1928 In ooples/AiDotNet;

· ooples opened on Jul 29, 2026

Audit: 54 AiModelBuilder Configure* methods silently drop their configuration

Status: Open.

#1876 In ooples/AiDotNet;

· ooples opened on Jul 15, 2026

Add finite-difference gradcheck as a model-family test invariant (replace convergence-only validation)

Status: Open.

#1872 In ooples/AiDotNet;

· ooples opened on Jul 13, 2026

PR #1789: GraphAttentionNetwork training is nondeterministic (flaky in isolation, not a CI flake)

Status: Open.

#1860 In ooples/AiDotNet;

· ooples opened on Jul 12, 2026

[Tracking] PR #1789 — green all remaining CI shards

Status: Open.

#1859 In ooples/AiDotNet;

· ooples opened on Jul 12, 2026

PR #1789: split Generated Layers G-M and N-S shards (45-min timeout)

Status: Open.

#1858 In ooples/AiDotNet;

· ooples opened on Jul 12, 2026

PR #1789: Finance CSDI/TFC forecasting-foundation — reshape + ParallelForOrSerial OOR

Status: Open.

#1856 In ooples/AiDotNet;

· ooples opened on Jul 12, 2026

PR #1789: CI parallel-execution flakes (GAT, SpiralNet, XDecoder, CpuToWallRatio, MemoryBudget)

Status: Open.

#1857 In ooples/AiDotNet;

· ooples opened on Jul 12, 2026

PR #1789: XLMRoBERTaNER MoreData diverges (loss 2.84 -> 9.40)

Status: Open.

#1855 In ooples/AiDotNet;

· ooples opened on Jul 12, 2026

PR #1789: DBNet training diverges (loss 0.21 -> 1.48)

Status: Open.

#1854 In ooples/AiDotNet;

· ooples opened on Jul 12, 2026