Skill UI

Orchestrates PyTorch-native TorchTitan 4D parallelism (FSDP2, TP, PP, CP) for large-scale LLM pretraining, scaling from 8 to 512+ GPUs with Float8, torch.compile, and distributed checkpointing to boot Llama 3.1, DeepSeek V3, or custom models.

A comprehensive guide demonstrating how to generate your first AI image using the Ideogram API. It provides runnable examples in cURL, TypeScript, and Python, covering both legacy and V3 endpoints. Users will learn essential API parameters, such as prompt structuring, style types, and aspect ratios, making it ideal for initial integration testing and setup.

This skill provides comprehensive guidance on training Mixture of Experts (MoE) models, including architectures like Mixtral and DeepSeek-V3. It details techniques for achieving large-scale model capacity with significantly reduced computational costs (e.g., 5x reduction vs. dense models). Topics covered include top-k routing, load balancing, expert parallelism, and integration with frameworks like DeepSpeed and HuggingFace for efficient, resource-constrained training.

This guide provides comprehensive techniques for optimizing ElevenLabs Text-to-Speech (TTS) performance. Learn how to reduce latency and maximize throughput by strategically selecting models (e.g., Flash vs. v3), implementing streaming endpoints (HTTP or WebSocket), and choosing optimal audio formats. It is essential for real-time conversational AI, IVR systems, and high-volume content generation.