1.2K skills · sorted by relevance
Build Python workflows that constrain language model output using text patterns and grammar rules. Generate structured JSON, restrict classifications to allowed labels, extract information, and combine reusable generation steps with loops and tool selection.
Helps fine-tune large language models with LoRA and QLoRA, methods that train small sets of additional parameters. Covers memory-conscious training, configuration choices, saving adapters, and switching task-specific adapters on a shared base model.
Guides experiments that change internal values in PyTorch models to test what causes their behavior. Uses pyvene to swap or remove activations, locate influential components, train interventions, and save experiments for reuse.
Clean, combine, summarize, and reshape tabular data with pandas DataFrames, Python's table-like data structures. Produces transformation code and validation checks, including time series processing and memory-conscious approaches for large datasets.
Evaluate Hugging Face Hub models using inspect-ai or lighteval, with local inference or a provider-backed option. Choose an inference engine, start with a small test, and switch engines when model compatibility requires it.
Fine-tune vision models on custom images using Hugging Face Jobs cloud GPUs. Covers object detection, image classification, and segmentation—identifying image regions—with dataset validation, hardware selection, cost estimates, monitoring, and saving trained models to the Hub.
Choose a serving container for models deployed on SageMaker. Match the model type to a compatible image, retrieve its address from the AWS catalog, and check regional availability, CPU or GPU variants, and required host settings.
Update the data structure of a deployed Convex app through staged changes. Using @convex-dev/migrations, the skill fills or transforms existing records, checks validity and record counts, and makes new fields required after migration finishes.
Process Excel workbooks by saving Python scripts to files and running them through the shell. Covers multiple sheets, data transformations, formulas, formatting, and error reporting for complex spreadsheet tasks.
Process Excel workbooks and generate PDF or Word documents by running Python scripts directly through run_shell. The skill covers multi-sheet edits, formatting, data transformations, document layout, and error handling for complex tasks that can exceed agent step limits.
Process Excel files with Python after checking that input files and source URLs are accessible. The workflow covers data transformations, formulas, and formatting, with alternative sources and error logs when access fails.
Process Excel workbooks and generate PDF reports by running Python scripts directly. Covers multiple sheets, data transformations, formatting, and error handling, with scripts saved as files for complex tasks.
Validate spreadsheet sources before running Python scripts to transform data, update multiple sheets, or apply formatting. Check required data first, try backup sources when needed, and report errors clearly if processing cannot proceed.
Read, write, and transform workspace files by running Python through run_shell when execute_code_sandbox cannot access them. The workflow covers explicit file paths, batch processing, error handling, and checking generated outputs.
Continue file reading and data processing when read_file or execute_code_sandbox fails. The skill uses run_shell with short Python commands or multiline scripts to inspect files, parse data, and save transformed output.
Write GSAP animations for HyperFrames compositions, from element entrances to coordinated sequences. Use timelines to organize timing, control playback, adapt motion to screen size and reduced-motion preferences, and choose animation properties suited to efficient rendering.
Transform an article, case study, or brand story into a connected scrolling page. Choose scroll-controlled video, a real-time Three.js scene, or HTML with charts and typography, then organize the narrative into five to seven clear stages.
Build visual sequences that move forward and backward with page scrolling. Use video, image frames, canvas, SVG, or HTML elements in a sticky stage, while preserving keyboard access, static alternatives, and predictable visual states.
Connect streaming APIs to Ant Design X by implementing a Chat Provider, an adapter for request and message formats. Choose built-in providers where suitable, transform custom responses, and configure authentication, retries, and stream parsing.
Connect agent tasks into a sequence where each step receives the previous output. Use custom prompts or predefined workflows to analyze code, plan refactoring, generate tests, optimize performance, or transform data through multiple stages.
This skill should be used when working with pre-trained transformer models for natural language processing, computer vision, audio, or multimodal tasks. Use for text generation, classification, question answering, translation, summarization, image classification, object detection, speech recognition, and fine-tuning models on custom datasets.
The English summary is being prepared. Showing the original description.
Convert existing Hugging Face Transformers Trainer or TRL SFTTrainer training code into an NVFLARE federated job using flare.patch(trainer), local validation, and job export; use when the user names Hugging Face or preliminary source inspection identifies one Hugging Face owner, and not for manual PyTorch loops, Lightning, inference-only pipelines, deployment, or experiment workflows.
The English summary is being prepared. Showing the original description.
Balance, normalize, and transform Hi-C contact matrices using cooler and cooltools. Apply iterative correction (ICE), compute expected values, and generate observed/expected matrices. Use when normalizing or transforming Hi-C matrices.
The English summary is being prepared. Showing the original description.
Transform, clean, reshape, and preprocess data using pandas and numpy. Works with ANY LLM provider (GPT, Gemini, Claude, etc.).
The English summary is being prepared. Showing the original description.