Results for “cross-training”

Train or fine-tune sentence-transformers models for retrieval, similarity, clustering, classification, and reranking, with support for bi-encoders, cross-encoders, and sparse encoders.

Train and fine-tune transformer language models using TRL (Transformers Reinforcement Learning) with support for SFT, DPO, GRPO, KTO, RLOO, and reward model training via CLI commands.

Fine-tune and align language models using reinforcement learning with TRL, including SFT, DPO, PPO, GRPO, and reward model training.

Provides practical guidance for training Mixture-of-Experts Vision-Language Models in Megatron Bridge, comparing FSDP and 3D-parallel approaches with lessons from recent multimodal experiments.

Train or fine-tune language and vision models using TRL or Unsloth on Hugging Face Jobs cloud infrastructure, with support for SFT, DPO, GRPO, and reward modeling, plus GGUF conversion for local deployment.

Validate and configure packed sequences and long-context training in Megatron-Bridge, distinguishing offline packed SFT for LLMs from in-batch packing for VLMs with correct context parallelism constraints.

Train, evaluate, export, and run inference for Masked Auto-Encoder (MAE) models for self-supervised pretraining and fine-tuning of visual representations.

Run Megatron-LM (MLM) and Megatron Bridge training with mock or real data, covering correlation testing, available recipes, and multi-GPU examples.

Train, evaluate, export, and run inference for pose classification models using ST-GCN on skeleton keypoint sequences.

Repair and verify cross-references between SKILL.md files after a skill is created, renamed, removed, or restructured. Ensures every skill that calls another skill references the correct name, and every skill that is called has accurate "Called by" context. Load after universal-skill-creator creates a skill, after improve-skills completes a cycle, after a skill is renamed or removed, or when the user asks to fix cross-references, sync skill links, repair broken skill references, update skill cross-links, check skill references, or "are cross-links correct".

Mandatory training archive with model gating (APPROVED/REVIEW/DROP). Trigger when: (1) training run completes, (2) need to decide which models to deploy, (3) want historical training reference, (4) need checkpoint recommendations for overfitting.

Create exercise directory structures with sections, problems, solutions, and explainers that pass linting. Use when user wants to scaffold exercises, create exercise stubs, or set up a new course section.

Fix PPO training early-stop issues. Trigger when: (1) impossible drawdown values (>100%), (2) training stops too early, (3) need adaptive recovery instead of hard stop.

Coordinates purple team exercises by running MITRE ATT&CK-mapped attack scenarios with real-time detection testing and collaborative gap remediation.

Cross-model benchmark for gstack skills. (gstack)

Create exercise directory structures with sections, problems, solutions, and explainers that pass linting. Use when user wants to scaffold exercises, create exercise stubs, or set up a new course section.

Create exercise directory structures with sections, problems, solutions, and explainers that pass linting, then commit with git.

Facilitates discussion-based tabletop exercises for SOC teams to test incident response procedures, communication workflows, and decision-making under pressure without impacting production systems.

Post-train LLMs with reinforcement learning using the slime framework, which integrates Megatron-LM for training and SGLang for rollout generation.

Curriculum Learning

EVA-CLIP: Improved Training Techniques for CLIP at Scale

Post-execution skill crystallization and learning engine updates for WinDAGs. Runs after successful execution to update Thompson sampling parameters, track method quality, detect monster-barring, log near-miss events, and signal Kuhnian crises. Activate on "curator", "learning update", "skill crystallization", "Thompson sampling", "monster-barring", "near-miss", "Kuhnian crisis", "post-execution learning". NOT for pre-execution risk scanning (use windags-premortem), retrospective analysis (use windags-looking-back), or DAG construction (use windags-architect).

Cross-validation only estimates generalization if the split mimics the gap between

Plan and execute authorized red team engagements with structured attack path analysis, MITRE ATT&CK kill-chain planning, technique scoring, choke point identification, OPSEC risk assessment, and crown jewel targeting.

Rehearse

许三多（军旅剧虚构）认知与表达框架（压缩蒸馏）：钝感坚持、不抛弃不放弃、草根成长 触发：士兵突击 等。禁止军事违法教程

Expert guidance for implementing GRPO/RL fine-tuning with TRL for reasoning and task-specific model training.

Log workouts, plan routines, track progress, and get intelligent coaching for any fitness level.

Compares two or more band-tagged frameworks and produces a framework-neutral topic matrix showing coverage and gaps across all inputs, plus an optional reference-centric PLC crosswalk document when a reference framework is supplied.

Expert guidance for GRPO/RL fine-tuning with TRL for reasoning and task-specific model training

Agent validation v4.3.0 — Make agents act effectively by disabling harmful actions, lowering gates, and injecting cross-run learning

Use this skill to classify tabular data end-to-end in MATLAB — load a dataset, prepare and clean it, select promising classifiers, train them, and compare accuracies with cross-validation, holdout, or hyperparameter optimization plus statistical tests. TRIGGER when: user asks to classify tabular data, pick classifiers for a dataset, compare classifier accuracy, run cross-validation or a holdout evaluation, or find the best model with statistical uncertainty. DO NOT TRIGGER when: user has non-tabular inputs (images, sequences, time series), wants a regression model, is training a specific neural network architecture (use matlab-train-network), or wants cost-sensitive learning or an arbitrary class-prior vector (this skill only supports the built-in uniform-prior toggle for imbalanced data).

Track ML experiments with automatic logging, visualize training in real-time, optimize hyperparameters with sweeps, and manage model registry with W&B.