Tengjiao_codex_skill

Use when evaluating generative models — choosing metrics (FID, IS, KID, sFID, FDD, FVD, PRDC, LPIPS, SSIM, AuthPct, Vendi), setting up online or offline evaluation, feature extractor selection, distributed computation, memory management during sampling.

Use when optimizing PyTorch training speed or memory on CUDA GPUs — global flags, torch.compile, fused optimizers, mixed precision, gradient checkpointing, kernel fusion, memory layout, or latent-space training. Applies to any PyTorch training workload.

Use when curating, uploading, or managing HuggingFace datasets for ML training, including offline caching, preflight verification, and data directory conventions.

Use when structuring ML experiment configs with Hydra, adding new config groups, or debugging config resolution. Applies to any project using Hydra for hyperparameter management.

Use when designing ablation studies to compare model components, loss functions, or architectural choices. Covers synthetic data experiments, variant loops, production metrics, and W&B grouping.

Use when integrating W&B experiment tracking into ML training pipelines, including logging strategy, run configuration, and online/offline mode management.

Use when streaming large datasets from tar shards with WebDataset, replacing file-based DataLoaders, or precomputing encoder latents into shards.

Manage ts1v23 Slurm GPU availability on Iridis by checking active allocations, deciding whether remaining walltime is sufficient for training/debug, and submitting renewal jobs with gpu-jupyter.sh (H100), gpu-jupyter2.sh (A100), gpu-jupyter_ecs_a100.sh (swarm_a100 A100), gpu-jupyter3.sh (quad_h200 H200), or gpu-jupyter_dualh200.sh (dual_h200 H200). Use when the user asks to check current GPU resources, renew cards due to low remaining time, or run debug commands on allocated compute nodes.

Read an uploaded deep-learning research plan file and prepare all files needed to run the Codex DL long-running harness. Use when the user asks to convert a plan into runnable scaffold files, generate required file manifests, or regenerate the in-project mechanism document.

Sync one or more local Codex skill folders from this server to a GitHub repository with incremental commit and push. Use when user asks to upload/update skills, keep remote skill repo in sync, or troubleshoot skill publishing issues (SSH auth, git identity, remote URL rewrite).

Create or revise technical Markdown documents that need Mermaid diagrams, flow validation, and source-code evidence links. Use when docs describe workflows/dependencies/branches and should include clickable file:line references plus concrete code snippets.