aqDeveloper environment for foundation models

aq changelog

  1. aq v0.0.4

    Fitting every machine

    New

    • Machine-aware compute: probes VRAM; VLMs no longer hardcode float32. Auto bf16/fp16, batch, gradient checkpointing; OOM auto-retry shrink.
    • Windows install: the installer now works cleanly on Windows (Git Bash and aq.cmd).
    • aq update: reinstall the latest release (same entrypoint as curl …/install.sh | bash).

    Fixes

    • VLM vision towers: vision: accepts full ViT sizes (vit-h/14, ViT-H-14, …) and optional OpenCLIP weights via vision_pretrained / pretrained.
    • aq serve for every method: generation on VLM (LLaVA/Flamingo/CLIP), vision CNN/ViT, and tabular; --image / serve.image for multimodal. LLM methods unchanged.
  2. aq v0.0.3

    Teaching models to see

    New

    • Vision CNNs: family: vision / method: cnn with aq-owned LeNet, AlexNet, VGG, ResNet, Inception, EfficientNet, ConvNeXt. ImageFolder or path+label table. Train, eval, inspect.
    • Vision transformers: method: vit with aq-owned ViT, Swin, DeiT (CNN teacher), BEiT (VQ + block MIM → classify).
    • Vision-language: family: vlm: CLIP / SigLIP (aq dual encoders, recall@1); LLaVA / GPT-4V-style (projector + causal LM); Flamingo (perceiver + gated cross-attn).
    • Train TUI: live dashboard on a real terminal (bars, stats, step table). Pipes/CI stay plain tables. AQ_TUI=0 turns it off.
    • Eval TUI: scoreboard of its own (metric / score / verdict / probes), not the train monitor.

    Fixes

    • speculative: true: assisted decode at serve with draft_model:.
    • formats: true: real weight export after train (HF + GGUF; GPTQ/AWQ on CUDA; EXL2 when convert.py is set). Prefer formats: [gguf] on Mac.
  3. aq v0.0.2

    Refining and accelerating

    New

    • aq plot: charts for loss, jobs, and runs. Ask the agent for a graph, or run aq plot yourself.
    • aq spawn --kill: optional critic agent that tries to break your idea in the background.

    Cleaner train folder

    • Fewer folders on aq init. Use tools/ for scripts and custom training code, jobs/plans/ for cron and sweeps.
    • Agent memory moved to ~/.aq/memory/ (not inside the train). Old memory/ folders migrate automatically.
    • instructions.md is now experiment.md. Old files rename themselves the first time you run aq status or aq train.
    • Dropped unused slots: connections/, methods/, sandbox/, train.ts.
    • aq schedule still works: it runs aq job plan.

    Agent

    • After aq init, tell the agent the new folder name: it won't lose track.
    • The agent waits for you to say go before training or eval.

    Install

    • Re-running the install script cleanly replaces the old copy (fixes upgrade from 0.0.1).

    Training fixes

    • CatBoost and LightGBM work out of the box.
    • Llama/GPT transformer trains no longer fail on missing pad tokens.
    • MLM works on causal models like Llama, not just BERT.
    • Multi-token prediction (mtp) actually trains extra heads now.
    • paged_kv and speculative no longer block training.
  4. aq v0.0.1

    Building a developer environment for ML

    Concept

    aq is the developer environment for foundation models: frameworks, libraries, tools, and autonomous agents that help you design, test, and refine models. Built for precision and intent. Think Next.js for JS, but for ML: architecture, experiments, and building AI the way it should be built, deliberately, not by guessing.

    Trains live on disk. A train is a folder with your data, recipe, and results. aq is the CLI that runs them: train, eval, queue jobs, and chat with an agent that works inside that folder, with the same precision as writing code.

    Install

    curl -fsSL https://aq.aquin.app/framework/install.sh | bash

    Needs Node 18+, Python 3, and npm. After install: aq help and aq doctor .

    New in 0.0.1

    Start a project

    • aq init: new experiment folder (aq-experiment, or -new1, -new2, … if that name exists)
    • aq init my-project: same, with your name
    • aq fork: copy an experiment (skips old job logs and checkpoints)

    Train & evaluate

    • aq train: fit from recipe.yaml
    • aq eval: score against files in evals/
    • aq serve: try generation from a checkpoint
    • Tabular ML: linear/logistic regression, ridge, lasso, elastic net, trees, forests, boosting, Gaussian processes
    • LLM work: pretrain, SFT, full fine-tune, LoRA, QLoRA (NVIDIA + 4-bit), masked LM, span corruption, fill-in-the-middle, continued training from a checkpoint
    • Transformers: encoder, decoder, encoder-decoder (set model: in the recipe)

    Work in the background

    • aq job run: run a command as a queued job
    • aq job list | log | cancel | resume | tree

    Agent & chat

    • aq: interactive chat in the terminal
    • aq ask: one-shot question, no UI
    • aq chat list: pick up old conversations (stored under ~/.aq/chats/)
    • aq spawn: start a background agent on a task
    • Plug in your own scripts (tools/) and skill packs (skills/)

    Other

    • aq provider: OpenAI, Anthropic, Grok, or Ollama
    • aq schedule: chains like train → eval, sweeps, cron-style reruns
    • aq stage: sub-experiments inside stages/
    • aq status, aq diff, aq checkout: see what ran and go back to an older state
    • aq version, aq doctor

    Fixes in 0.0.1

    • Install: aq command works after curl install (no broken path / missing binary)
    • Install: Python deps on Ubuntu/Debian (venv + pip setup)
    • Training: faster LM runs on typical GPUs; less likely to OOM on a T4-class card
    • Chat: conversations live in one global place, not scattered per folder

    This is only the first cut. We want aq to feel like a calm desk for teaching models: yours to keep, easy to share, honest about what it can do. If you are curious about how machines learn, you already belong here. Build something small. See what it becomes.

Not sure if Aquin is right for you?