Directory
Published directory entries matching your search.
A Hugging Face Space that accepts uploaded or recorded audio and returns a text transcription with grammar correction. The page reports a runtime error because none of PyTorch, TensorFlow >= 2.0, or Flax were found, so models are unavailable.
StableLM: Stability AI Language Models.
An MLOps/LLMOps platform for model building, evaluation, and fine-tuning.
StreamingLLM is a technique from researchers at Meta AI, MIT and Carnegie Mellon University that extends a language model's context to millions of tokens without large compute or memory resources. The article describes it as addressing the fixed context length of models such as Llama-2, which has a context of approximately 4,000 tokens.
Open-source project that supports AI agents, automated workflows, orchestration, or delegated tasks.
A neuro-symbolic framework for building applications with LLMs at the core.
Inference engine for TensorRT on Nvidia GPUs.
TextGen is an open-source desktop app for local LLMs, with text, vision, tool-calling and web search, plus a UI and API. It supports multiple backends including llama.cpp, Transformers, ExLlamaV3 and TensorRT-LLM, and an OpenAI/Anthropic-compatible API. It also offers LoRA training and image generation. Portable builds are provided for Linux, Windows and macOS.
The Hundred-Page Language Models Book by Andriy Burkov is a concise hands-on guide covering machine learning basics through advanced transformer architectures and large language models. It explains AI fundamentals, text representation, count-based models, recurrent neural networks, and transformers, with mathematical foundations and Python code examples. The book targets ML practitioners and engineers building text-based applications.
An essay by Artur Piszek that describes the AIKEA Effect, analogous to the IKEA Effect, as applied to AI products. It argues that current models such as GPT-3.5, GPT-4 and LLAMA2 work well out of the box and advises against overengineering when working with LLMs. It outlines low-hanging AI use cases and the tendency of experienced engineers to build unnecessary complexity.
A repository collecting system prompts, custom instructions and jailbreak prompts for various LLM providers and solutions, including ChatGPT, Microsoft Copilot, Claude, Gab.ai, Gemini and Cohere. It also contains articles, tools, and scripts, and states that its content is intended for learning and informational purposes.
The Dark Risk of Large Language Models WIRED UK provides machine-learning models, research, training resources, or evaluation tools.
The Pile is an 825 GiB open source language modelling data set made of 22 smaller high-quality datasets combined together. It is offered as a training set and as a benchmark, and is distributed as jsonlines data compressed with zstandard.
A visualisation of major large-language models ranked by performance using the MMLU benchmark, which consists of 16,000 multiple-choice questions across 57 academic subjects. It also presents selected lawsuits filed against AI companies and notes that MMLU has some critiques.
Agent techniques to augment your LLM and push it beyond its limits.
TrueFoundry is an enterprise AI gateway that sits between applications and LLM providers and MCP servers. It offers a unified API for accessing more than 1000 LLMs, along with routing, fallbacks, rate limiting, access control, budgets, guardrails and observability. Deployment options include SaaS, hybrid and self-hosted VPC environments.
uithub provides GitHub repository context for use with LLMs. Changing one letter in a GitHub URL, from G to U, returns full repository context. The site also offers an API for obtaining LLM context and lists example uses such as code verification, code search and changelog tracking.
Unsloth is a desktop application for running and training LLMs and diffusion models locally. It supports model formats including GGUF and MLX, and offers fine-tuning methods such as LoRA, QLoRA, full fine tuning and reinforcement learning, along with dataset building and model export. Native builds are offered for Windows, macOS and Linux, with additional installation options.
Any URL to clean markdown for LLMs. Free API, no signup required. Strips JS/CSS and outputs LLM-ready content.
Valohai is an MLOps platform for AI products that combine LLMs and specialized models. It provides experiment tracking, data versioning, hyperparameter optimization, pipeline automation and compute orchestration, and runs on existing AWS, Azure, GCP or on-prem infrastructure. Pricing is per user license across Pro and Enterprise tiers.
Voyager is an LLM-powered embodied agent for Minecraft that explores, acquires skills, and makes discoveries without human intervention. It combines an automatic curriculum, a skill library of executable code, and an iterative prompting mechanism using environment feedback, execution errors, and self-verification. It interacts with GPT-4 through blackbox queries, avoiding model parameter fine-tuning.
Weco is a research and product lab building recursively self-improving AI. Its platform points at an eval script and iteratively optimizes code against a metric the user defines, such as speed, accuracy or cost. Supported model providers include GPT, Claude, Gemini, Kimi, GLM and MiniMax, with bring your own key.
Documentation for Weights & Biases (Prompts), covering setup, features, and practical usage.
WizardLM is a repository of large language models including WizardLM, WizardCoder and WizardMath, described as LLMs built upon Evol Instruct. The page lists model checkpoints and benchmark results, with WizardCoder-33B-V1.1 reported at 79.9 pass@1 on HumanEval and WizardMath-7B-V1.1 reported at 83.2 pass@1 on GSM8k.
