Skip to main content

LLM

perplexity.ai chatgpt gemini copilot claude.ai deepseek

Voice / Speech​

Phone Call > LLM Into A Human-Like Voice Agent https://www.youtube.com/watch?v=0LT64_mgkro

https://github.com/PennyroyalTea/gibberlink

https://www.forbes.com/sites/dianehamilton/2025/02/25/what-is-gibberlink-mode-ais-secret-language-and-way-of-communicating/ vLLM

  • What it is: vLLM is an open-source library for efficient inference and serving of large language models (LLMs).
  • Key Features:
    • High Throughput: vLLM significantly improves the speed of running LLMs compared to traditional libraries like Hugging Face Transformers. It achieves this through optimizations like PagedAttention, which efficiently manages memory usage.
    • Compatibility: Works with a wide range of popular LLMs, including those from Hugging Face.
    • Ease of Use: Relatively easy to install and use, making it accessible to developers.
    • Open Source: Allows for community contributions and customization.

Hugging Face

  • What it is: Hugging Face is a leading platform for building, training, and deploying machine learning models, with a strong focus on natural language processing (NLP).
  • Key Aspects:
    • Model Hub: A vast repository of pre-trained models, including many popular LLMs, datasets, and other NLP resources.
    • Transformers Library: A powerful and popular library for working with transformer-based models, the architecture underlying many LLMs.
    • Community: A large and active community of developers and researchers contributing to and utilizing the platform.
    • Ease of Use: Provides user-friendly tools and APIs for interacting with models and datasets.