training-llms-megatron

Streamlines AI research workflows by providing curated Claude skills for data analysis, literature review, experiment design, and research paper generation.

Assists AI researchers in drafting, structuring, and generating machine learning research papers with academic writing best practices and technical precision.

Streamlines distributed machine learning training using Ray, optimizing hyperparameter tuning and parallel model execution across compute clusters.

Streamlines distributed data processing and machine learning workflows using Ray's scalable data loading and transformation capabilities.

Generates structured document outlines and hierarchical content maps with customizable depth and formatting for research and writing workflows

Provides structured, context-aware advice and recommendations for complex problem-solving, research workflows, and strategic decision-making

Streamlines parameter-efficient fine-tuning of large language models using Transformers Reinforcement Learning (TRL) techniques and best practices.

Automates seamless deployment and management of AI workloads across multiple cloud providers with intelligent resource optimization and cost-efficiency.

Trains large language models using open-source reinforcement learning from human feedback (RLHF) techniques with advanced alignment and reward modeling

Validates and sanitizes AI prompts to prevent injection attacks, filter sensitive content, and ensure safe, controlled interactions with language models.