Data analysis, machine learning, LLM apps, and prompt design (9K skills)
Compress LLM weights using HQQ quantization, which stores weights with fewer bits without calibration datasets. Follow workflows for precision settings, model saving, inference integration, and LoRA fine-tuning of quantized models.
Guides language model training with GRPO in TRL, comparing several responses and rewarding better results. Covers dataset preparation, scoring rules for correctness and format, training configuration, monitoring, and saving models for later use.
Guide the design of AI product features with output validation, responses displayed as they arrive, prompt versioning, and testing. Address API costs, failures, and unsafe input when preparing language model integrations for production.
Guides large language model reinforcement learning with miles, covering environment setup, low-precision training, consistency between training and inference, and speculative decoding, where a smaller model proposes tokens for a larger model to verify.
Use Outlines to make language models produce data in a defined format. The skill covers extracting information into JSON, restricting classification labels, and applying Pydantic models, which define fields and data types, with local model backends.
Plan and configure large language model training with OpenRLHF to better match human preferences. Covers training method selection, distributed GPU setup, inference acceleration with vLLM, and troubleshooting memory or training stability issues.
Provides prompt templates for coding, writing, analysis, and creative work. Helps define the AI's role, specify the desired output, add examples and constraints, and refine requests using a practical checklist.
Build searches with Qdrant that find similar content and filter results by attributes. The skill covers document retrieval for AI answers, vector storage, memory optimization, and distributed deployment for larger collections.
Design and refine prompts for language model applications. Choose useful examples, create reusable templates and system instructions, and compare prompt variations using accuracy, consistency, response time, and token usage to guide improvements.
Guides Python agent development with PydanticAI, covering tool calls and structured outputs checked against defined data types. Explains service injection, provider switching, streaming responses, and tests that run without calling a real language model.
Design RAG systems that retrieve relevant documents to support LLM answers. The skill guides document splitting, meaning-based and keyword search, result reranking, and separate retrieval evaluation to improve the context passed to the model.
Design and improve RAG systems, which retrieve documents to support language model answers. Explore meaningful document splitting, vector storage, combined keyword and semantic search, and relevance ranking while checking common causes of poor retrieval.
Reduce the memory needed to run large language models with GPTQ quantization, which stores model weights at lower precision. Covers loading compressed models, quantizing your own models, choosing execution backends, and connecting to memory-efficient fine-tuning.
Improve AI instructions with examples, clear output formats, and comparison tests. Use reusable templates and system prompts, which set ongoing behavior, to address inconsistent answers and investigate unexpected agent behavior.
Guide language model training with TRL, from learning to follow instructions to matching human preferences. Covers method selection, training data, reward models that score responses, evaluation, and adjustments for memory limits or unstable training.
Explore training workflows that reduce harmful model responses using a written set of principles. The skill covers self-critique, revised training examples, and RLAIF, reinforcement learning from AI feedback, with implementation examples and hardware guidance.
Build search and recommendation features with Pinecone, a managed database for numerical representations of content. Covers data loading, combined meaning and keyword search, filtering by attributes, and separating data by user.
Guide reinforcement learning, training models through rewards, with slime. Prepare data, configure Megatron-LM training and SGLang response generation, and set up asynchronous runs or custom multi-turn workflows for supported language models.
Generate embeddings, numerical representations of text or images, for meaning-based search, similarity comparisons, and document grouping. The skill guides model selection, multilingual processing, local execution, and integration with retrieval workflows.
Build similarity search with FAISS to find nearby matches in large collections of numeric vectors. The skill guides index selection, memory reduction, GPU use, and saving indexes while balancing search speed and accuracy.
Plan caching for repeated LLM requests by choosing what to reuse: shared prompt beginnings, complete responses, or documents loaded in advance. Covers Anthropic prompt caching, cache invalidation, and problems caused by changing inputs or outdated answers.
Build workflows that extract structured data from LLM responses, check field types and values with Pydantic, and retry failed extractions. Includes examples for nested data, classification, and receiving results as they are generated.
Build image classification and search workflows by comparing images with text descriptions. Covers classification without additional training, similarity scoring, batch processing, and storing image representations in a vector database for retrieval.
Guides experiments that inspect and change internal model values using nnsight. Covers local PyTorch models and remote experiments through NDIF, including replacing intermediate values to test how particular layers affect a model’s predictions.