nanochat

A minimal-code repository for training a small GPT-2-style chat model.

LibraryAI & MLllm-traininggpt-2minimaleducation

Recommended in 1 thread

Attention Is Not What You Need (2025)
arxiv.org · 17 hours ago
Recommended for hands-on experiments with architectural tweaks.

chatjimmy ↗AI & ML
A chatbot service known for extremely fast text generation.
1 thread
FUTO ASR4All ↗AI & ML
FUTO's small CPU-optimised streaming English ASR models.
1 thread
clojure-mcp AI & ML
MCP tooling that lets LLM agents edit Clojure code and interact with a REPL.
1 thread
TabPFN AI & ML
A transformer model that does tabular prediction via in-context learning.
1 thread
DeepSeek V4.1 Flash AI & ML
A low-cost DeepSeek language model available via API.
1 thread
Pydantic AI AI & ML
Python agent framework with abstractions over LLM providers.
1 thread
Liquid AI Decision Models ↗AI & ML
Models from Liquid AI trained specifically to produce decisions with calibrated confidence.
1 thread
OpenCode Go ↗AI & ML
A subscription plan for accessing models through the OpenCode coding tool.
2 threads

OpenRouter ↗
A marketplace that routes API requests to many LLM inference providers.
7 threads
GLM ↗
An open-weights large language model family from Zhipu AI.
5 threads
llama.cpp
A C/C++ engine for running LLMs locally.
4 threads
GoModel ↗
LLM gateway written in Go, offered as a LiteLLM alternative.
3 threads
DeepSeek
A family of large language models offered via a low-cost API.
2 threads
OpenCode Go ↗
A subscription plan for accessing models through the OpenCode coding tool.
2 threads
Cerebras ↗
A hosted fast inference service offering Qwen models.
1 thread
Claude
Anthropic's AI model and command-line assistant.
1 thread
Claude Code gateways ↗
Documented setup for routing Claude Code through an LLM gateway for centralized control and alternate model providers.
1 thread
Cloudflare Workers AI ↗
Cloudflare's serverless platform for running hosted AI models.
1 thread
CodeCarbon ↗
An open-source Python library for estimating the carbon emissions of compute workloads.
1 thread
Codex
OpenAI's coding agent.
1 thread