Discover ML and AI skills during model training and inference workflows.
npx skills add https://github.com/rand/cc-experiments --skill discover-ml
Please help me install this Agent Skill. Skill: discover-ml Source: https://github.com/rand/cc-experiments/tree/main/skills/discover-ml Command: npx skills add https://github.com/rand/cc-experiments --skill discover-ml
The rapidly evolving field of Machine Learning and AI, especially with LLMs and diffusion models, presents significant challenges in model selection, training, and evaluation. This skill provides instant access to expertise on RAG, LLM evaluation, fine-tuning, and multi-model orchestration, streamlining your AI development.
Help me evaluate the performance of my RAG system using appropriate metrics.
High-intent search queries and answers about installing and using this skill.
RAG evaluation requires measuring retrieval quality and generation accuracy. Use metrics like precision, recall, and BLEU scores for retrieved documents, combined with semantic similarity and factual correctness for generated outputs. This skill provides evaluation frameworks tailored to RAG pipelines.
Fine-tuning adapts pre-trained LLMs to domain-specific tasks by training on curated examples. The process involves preparing datasets, selecting optimizer settings, and monitoring loss curves. This skill covers fine-tuning workflows with PyTorch and practical evaluation strategies.
RAG combines document retrieval with LLM generation to ground responses in external data. Implementation steps include indexing documents, embedding queries, retrieving relevant passages, and feeding them to the LLM. This skill guides you through RAG architecture, component selection, and evaluation.
Multi-model orchestration coordinates different frameworks and model architectures in unified workflows. PyTorch and TensorFlow can coexist through abstraction layers and shared inference servers. This skill covers orchestration patterns, model selection, and framework integration.
LLM evaluation combines quantitative metrics—perplexity, ROUGE, BLEU—with qualitative assessment of factuality, coherence, and relevance. Custom evaluation frameworks may assess domain-specific criteria. This skill provides evaluation methodologies and metric selection for different LLM tasks.
Diffusion model evaluation measures image quality, diversity, and alignment with prompts using FID, CLIP scores, and human assessment. Framework setup involves data pipelines, metric computation, and result visualization. This skill covers diffusion evaluation foundations and deployment considerations.
luokai25Train and deploy deep learning models with PyTorch and TensorFlow.
ToXMonDetect data drift and monitor model health for Domino-deployed models.
OmidZamaniFine-tune a DSPy program into deployable model weights via BootstrapFinetune.
DarkArty07Coordinate RLHF and post-training alignment workflows for LLMs using TRL.
