72 skills · sorted by relevance
Guide knowledge distillation, where a smaller language model learns from a larger teacher. Covers training from teacher responses or prediction scores, adjusting training settings, and comparing outputs to support model compression and specialized training.
Choose what kind of knowledge to transfer between teacher and student models: response, feature, or relational, and decide among offline, online, self, or cross-modal distillation schemes. Best for distillation strategy selection, capacity-gap diagnosis, and transfer planning. Activate on "knowledge distillation", "teacher-student", "soft labels", "feature distillation", "online distillation", or "cross-modal transfer". NOT for generic compression checklists or unrelated training work.
The English summary is being prepared. Showing the original description.
Train smaller language models using knowledge distillation, where a larger model supplies examples or prediction targets. This skill covers training losses, generated training data, parameter choices, and output comparisons to support model compression and specialized student models.
Design deployment-focused distillation systems that balance model size, accuracy, calibration, and cascade escalation under real resource limits. Best for teacher-student compression, threshold design, and failure-aware deployment. Activate on "model compression", "teacher- student", "distillation score", "cascade model", "edge deployment", or "model calibration". NOT for generic deep-learning overviews, prompt optimization, or training work without a concrete distillation objective.
The English summary is being prepared. Showing the original description.
Compress large language models using knowledge distillation from teacher to student models. Use when deploying smaller models with retained performance, transferring GPT-4 capabilities to open-source models, or reducing inference costs. Covers temperature scaling, soft targets, reverse KLD, logit distillation, and MiniLLM training strategies.
The English summary is being prepared. Showing the original description.
Deep analysis of knowledge distillation techniques for compressing large models into smaller efficient ones
The English summary is being prepared. Showing the original description.
Comprima modelos de linguagem grandes usando destilação de conhecimento de modelos professor para aluno. Use ao implantar modelos menores com desempenho retido, transferir capacidades do GPT-4 para modelos de código aberto ou reduzir custos de inferência. Aborda escalamento de temperatura, alvos suaves, KLD reversa, destilação de logits e estratégias de treinamento MiniLLM.
The English summary is being prepared. Showing the original description.
SkillAlchemy — One thought conceived, one goal achieved. Accept any idea or distillation target and produce an installable SKILL.md. It uses Lens to clarify the problem and LEAP to run distillation or fusion. This is the sole user-facing entry point. Use when the user asks to distill, generate a skill, fuse skills, or says, "I want to build X, but I do not know where to start."
The English summary is being prepared. Showing the original description.
Expert-level separation processes covering distillation, absorption, extraction, crystallization, membrane separations, and adsorption for chemical engineering applications.
The English summary is being prepared. Showing the original description.
Use when the operator says "run distillation" or "refresh complexity corpus", or a distillation pass is due.
The English summary is being prepared. Showing the original description.
Set up and run knowledge distillation (on-policy, off-policy, or multi-teacher) from a teacher model to a student model using the Tinker API. Use when the user wants to distill knowledge, compress models, or train a student from a teacher.
The English summary is being prepared. Showing the original description.
Build training datasets for LLM specialization from production data, frontier model distillation, and synthetic bootstrapping. Use when formatting production logs into SFT data, distilling from frontier APIs, or preparing data for fine-tuning. Covers JSONL formatting, data quality validation, deduplication, and train/eval splitting.
The English summary is being prepared. Showing the original description.
Create and edit videos using a measured reference style for color and cutting rhythm. Combine generated clips, existing footage, overlays, and 3D props, then check visual measurements and deliver both a video and an editable timeline.
Measure the colors and shot lengths of reference videos to build a reusable style pack. Outputs include a LUT, a color conversion table, plus still frames, overlay assets, and a generation specification grounded in the measured footage.
Build agent memory with ReasoningBank and AgentDB. Record actions and outcomes, retrieve similar experiences, assess results, and consolidate memories into reusable patterns so agents can draw on past work when making decisions.
Build agent memory with ReasoningBank and AgentDB. Record actions and outcomes, retrieve similar experiences, assess results, and consolidate memories into reusable patterns so agents can draw on past work when making decisions.
Plan further training for AI agent models by diagnosing gaps in knowledge, output consistency, or decision making. Compare continued pretraining, supervised fine-tuning, reinforcement learning, and distillation, then design training data, feedback environments, and evaluation checks.
Synthesize OpenViking knowledge bases or document collections into topic-based Markdown pages that explain trends, changes, comparisons, and implications. Keep conclusions traceable to source evidence, distinguish facts from inferences, and preserve disagreements and uncertainty.
Train sentence-transformers models for search, similarity, and reranking, which reorders search results by relevance. Choose a model type, prepare data, configure training, compare results against a baseline, and publish the trained model to Hugging Face Hub.
Run batches of machine learning experiments on remote GPUs over SSH, varying settings and random seeds. The skill schedules jobs, tracks progress, retries GPU memory failures, and cleans up leftover sessions before starting subsequent batches.
Train sentence-transformers models for search, similarity, and reranking, which reorders search results by relevance. Choose a model type, prepare data, configure training, compare results against a baseline, and publish the trained model to Hugging Face Hub.
Build agent memory with ReasoningBank and AgentDB. Record actions and outcomes, retrieve similar experiences, assess results, and consolidate memories into reusable patterns so agents can draw on past work when making decisions.
[omh] Turn repeated lessons into written rules: extract repeated principles from skills, prompts, traces, reviews, and failures into reviewed rule candidates without auto-mutating guidance. Use when the user says: rules-distill, rules distill, distill rules, rule distillation, principle distill, skill principles, extract agent rules, turn traces into rules.
The English summary is being prepared. Showing the original description.
Fine-tune models on Microsoft Foundry using SFT (supervised), DPO (preference), or RFT (reinforcement with graders). Covers dataset preparation, training job submission, deployment, and evaluation. USE FOR: fine-tune, SFT, DPO, RFT, training data, grader, distillation, fine-tuned model, training job, large file upload, calibrate grader, deploy fine-tuned model, evaluate fine-tuned model. DO NOT USE FOR: general model deployment without fine-tuning (use deploy-model), agent creation (use agents), prompt optimization without training (use prompt-optimizer).
The English summary is being prepared. Showing the original description.