awesome-repositories.com
Blog
MCP
awesome-repositories.com

Discover the best open-source repositories with AI-powered search.

ExploreCurated searchesOpen-source alternativesSelf-hosted softwareBlogSitemap
ProjectMCP serverAboutHow we rankPress
LegalPrivacyTerms
© 2026 Bringes Technology SRL·VAT RO45896025·hello@awesome-repositories.com
arman-bd avatar

arman-bd/guppylm

0
View on GitHub↗

Guppylm

GuppyLM is a small educational project that trains a compact language model from scratch and gives it the voice of a fish. It covers the whole path end to end: generating a synthetic conversation dataset, training a tokenizer, training a decoder-only transformer, and chatting with the result. The aim is to make the full lifecycle of a language model easy to follow.

The training data comes from a template-based generator. It combines randomized components into a fixed set of templates, producing tens of thousands of consistent, character-driven question-and-answer samples across dozens of topics, with no manual writing. The tokenizer is learned directly from that generated text and ships alongside the model as a separate small file.

The trained model can be used in two ways. A local terminal session answers typed prompts interactively or replies to a single prompt and exits. The model's weights are also quantized to a reduced precision, which makes them small enough to run inside a web page using WebAssembly. The page downloads the weights and tokenizer when it loads and runs the model on the user's own device. No server, install, or API key is needed. The browser chat includes clickable preset prompts for trying the model quickly.

Hosted notebooks chain dataset download, tokenizer training, model training on a free GPU, and a chat test into one workflow. They also include chatting with a pre-trained model.

AI search

Explore more awesome repositories

Describe what you need in plain English — the AI ranks thousands of curated open-source projects by relevance.

Start searching with AI

Features

  • AI Chatbots - Runs a quantized language model chatbot entirely inside the web page using WebAssembly, needing no server, installation, or API key.
  • Language Model Training - Trains a compact decoder-only transformer language model from scratch, covering synthetic dataset generation, tokenizer training and model training in one end-to-end workflow.
  • Chat Model Interactions - Lets users send messages to a small language model in the browser and read replies written in the voice of a fish.
  • From-Scratch Decoder Implementations - Implements a compact decoder-only transformer built from scratch that predicts the next token, kept small so the whole lifecycle is easy to follow.
  • In-Browser Model Inference - Runs a quantized model entirely inside the web browser using WebAssembly, so no server, install, or API key is needed.
  • Small Language Models - Provides a small conversational language model that answers in the consistent voice of a fish, trained on synthetic question-and-answer dialogue.
  • Template-Based Generation - Builds conversation datasets by filling a fixed set of templates with randomly combined components, with no manual writing.
  • Synthetic Dataset Generators - Generates tens of thousands of consistent, character-driven question-and-answer training samples across many topics by combining randomized template components.
  • Weight Quantization - Stores model weights at reduced numeric precision so the download is small enough to fetch and run inside a web page.
  • Synthetic Data Generators - Generates tens of thousands of consistent, character-driven question-and-answer samples across dozens of topics from randomized template components.
  • Tokenizer Training - Learns a vocabulary directly from the generated conversation text and saves it as a small separate tokenizer file shipped with the model.
  • Training Notebooks - Provides hosted notebooks that chain dataset download, tokenizer training, model training on a free GPU, and a chat test in one workflow.
  • WebAssembly Inference Runtimes - Compiles the inference runtime to WebAssembly so the model runs inside the page on the user's own device without a server.
4,045 stars·381 forks·Python·12 views

Star history

Star history chart for arman-bd/guppylmStar history chart for arman-bd/guppylm

How this analysis was created: This summary and feature list are AI-generated from collected project material and can contain mistakes. Stars, license and language are imported from GitHub. Inclusion does not mean that we have tested or audited this project. Check the source documentation for any feature you depend on. Learn more on our About page.

Projects sharing features with Guppylm

These projects share indexed features with Guppylm. Shared tags can include platform or build tooling; verify the primary use case before treating a result as a replacement.
  • huggingface/smollmhuggingface avatar

    huggingface/smollm

    3,949View on GitHub↗

    SmolLM is a project dedicated to the development of small language models. It focuses on training and fine-tuning compact models that maintain high performance while utilizing fewer parameters. The project emphasizes efficient AI inference and on-device text generation, aiming to enable the deployment of lightweight models on edge devices with limited memory and processing power. It utilizes synthetic data generation to produce artificial datasets that improve the reasoning and training of these AI systems. The system supports a variety of optimization and training capabilities, including we

    Python
    View on GitHub↗3,949
  • meta-pytorch/torchtunemeta-pytorch avatar

    meta-pytorch/torchtune

    5,821View on GitHub↗

    Torchtune is a PyTorch-native library for fine-tuning, aligning, and quantizing large language models. It provides a config-driven system for instantiating components, orchestrating distributed training, and managing parameter-efficient fine-tuning with quantization support, all through YAML-based configurations and command-line overrides. The library distinguishes itself through its comprehensive post-training workflow orchestration, combining supervised fine-tuning, preference optimization (DPO, PPO, GRPO), knowledge distillation, and quantization-aware training in a single configurable pip

    Python
    View on GitHub↗5,821
  • zyds/transformers-codezyds avatar

    zyds/transformers-code

    3,782View on GitHub↗

    This project is a collection of scripts and workflows for training, fine-tuning, and deploying large language models using the Hugging Face Transformers toolkit. It functions as a distributed training framework, a library for natural language processing task implementations, and a system for building retrieval-augmented generation chatbots. The repository includes specialized tools for model optimization, such as a Bayesian hyperparameter optimizer for automatically tuning model settings. It provides implementations for scaling model training across multiple graphics processors using data par

    Jupyter Notebookhuggingfacepefttransformers
    View on GitHub↗3,782
  • karpathy/build-nanogptkarpathy avatar

    karpathy/build-nanogpt

    5,621View on GitHub↗

    This is an educational implementation that builds a generative pre-trained transformer (GPT) language model from scratch using PyTorch. The project is structured as a step-by-step tutorial, walking through the construction of a decoder-only transformer architecture and its training loop with clean git commits and an accompanying video lecture for a hands-on learning experience. What sets this implementation apart is its focus on practical reproduction: it provides a workflow to train a 124-million-parameter model from scratch in about one hour on cloud GPU hardware, costing under ten dollars.

    Python
    View on GitHub↗5,621
Compare all 30 related projects→

Frequently asked questions

What does arman-bd/guppylm do?

GuppyLM is a small educational project that trains a compact language model from scratch and gives it the voice of a fish. It covers the whole path end to end: generating a synthetic conversation dataset, training a tokenizer, training a decoder-only transformer, and chatting with the result. The aim is to make the full lifecycle of a language model easy to follow.

What are the main features of arman-bd/guppylm?

The main features of arman-bd/guppylm are: AI Chatbots, Language Model Training, Chat Model Interactions, From-Scratch Decoder Implementations, In-Browser Model Inference, Small Language Models, Template-Based Generation, Synthetic Dataset Generators.

Which projects share features with arman-bd/guppylm?

Projects with overlapping indexed features include: huggingface/smollm — SmolLM is a project dedicated to the development of small language models. It focuses on training and fine-tuning… meta-pytorch/torchtune — Torchtune is a PyTorch-native library for fine-tuning, aligning, and quantizing large language models. It provides a… zyds/transformers-code — This project is a collection of scripts and workflows for training, fine-tuning, and deploying large language models… karpathy/build-nanogpt — This is an educational implementation that builds a generative pre-trained transformer (GPT) language model from… intel-analytics/ipex-llm — ipex-llm is an acceleration library and inference engine designed to optimize the execution and finetuning of large… ml-explore/mlx-examples — This repository provides a collection of reference implementations and code examples for training and deploying…