Skip to content

Readme

Jump to bottom
Kai Piper edited this page Apr 18, 2026 · 4 revisions

Supermix

Supermix is the working monorepo for the current Supermix / ChampionNet / Omni Collective line.

This repository combines:

  • local-first chat and multimodal runtime code
  • experimental training and continuation pipelines
  • desktop EXE and installer packaging
  • benchmark tooling and graph generation
  • published-model export helpers
  • bundled datasets and generated research artifacts

It is intentionally a mixed workspace, not a minimal source-only model repo.

Current status

As of March 29, 2026:

  • the latest finished omni checkpoint in this repo is omni_collective_v4
  • the latest packaged desktop release is studio-desktop-20260329-omni-v4-allmodels
  • the installer bundle currently includes 23 zipped model artifacts from the local model-pack directory used by the desktop build
  • a v5 continuation path exists in source/ and is currently an in-progress local experiment, not a finished released model

What is in this repo

  • source/
    • active development workspace
    • training scripts, model definitions, dataset builders, benchmark runners, desktop packaging helpers
  • runtime_python/
    • packaged local runtime path
    • simpler run path than the full source/ workspace
  • datasets/
    • conversation, coding, reasoning, science, and related local training inputs
  • output/
    • generated artifacts, benchmark graphs, summaries, logs, Hugging Face upload folders
  • installer/
    • Inno Setup definitions for the desktop app
  • dist/
    • built EXEs and installer outputs
  • web_static/
    • lightweight browser-only metadata bundle

Main capabilities

  • multimodel desktop app with model selector, Auto routing, collective mode, and agent mode
  • local chat, image-prompt, math, science-image, and omni-fusion model families
  • native-image experimental checkpoints
  • training pipelines for frontier, omni, lite, and specialist model lines
  • benchmark sweeps across common text benchmarks
  • export and publishing workflows for GitHub releases and Hugging Face model/dataset repos

Quick start

Run the packaged runtime

python runtime_python/chat_web_app.py

Windows launchers:

runtime_python\launch_chat_web_supermix.bat
runtime_python\launch_chat_terminal_supermix.bat

Run the active source app

python source/chat_web_app.py

Run the desktop multimodel app from source

python source/supermix_multimodel_desktop_app.py

Run the browser-only static bundle

Open:

web_static/index.html

Current desktop release

Latest release published from this repo:

  • Release page:
    • https://github.com/kai9987kai/Supermix_29/releases/tag/studio-desktop-20260329-omni-v4-allmodels
  • Installer:
    • https://github.com/kai9987kai/Supermix_29/releases/download/studio-desktop-20260329-omni-v4-allmodels/SupermixStudioDesktopSetup.exe
  • EXE:
    • https://github.com/kai9987kai/Supermix_29/releases/download/studio-desktop-20260329-omni-v4-allmodels/SupermixStudioDesktop.exe

Local build outputs:

  • dist/SupermixStudioDesktop/SupermixStudioDesktop.exe
  • dist/installer/SupermixStudioDesktopSetup.exe
  • dist/installer/SupermixStudioDesktopReleaseSHA256.txt

Model families in the workspace

The repo contains code and artifacts for several model lines:

  • Qwen adapter line
    • v28
    • v30
  • Champion / frontier line
    • v31
    • v32
    • v33
    • v34
    • v35
    • v39
  • native-image line
    • v36
    • v37
    • v38
  • omni-collective line
    • v1
    • v2
    • v3
    • v4
  • specialist lines
    • math_equation_micro_v1
    • science_image_recognition_micro_v1

Latest finished omni model

The latest finished omni checkpoint in this repo is omni_collective_v4.

Key details from output/supermix_omni_collective_v4_frontier_20260329/omni_collective_v4_frontier_summary.json:

  • parameter count: 19,032,281
  • stage-1 rows: 8,589
  • stage-2 rows: 9,447
  • final stage-2 weighted validation score: 0.5176
  • final stage-2 validation:
    • intent: 0.8195
    • response: 0.1402
    • vision: 0.9020
    • domain: 0.7765

Local packaged artifact:

Hugging Face models

Public model repos already published from this workspace:

  • Kai9987kai/supermix-v33-frontier
  • Kai9987kai/supermix-omni-collective-v1
  • Kai9987kai/supermix-v38-native-image-xlite-fp16
  • Kai9987kai/supermix-v39-frontier-reasoning-plus
  • Kai9987kai/supermix-omni-collective-v2-frontier
  • Kai9987kai/supermix-math-equation-micro-v1
  • Kai9987kai/supermix-omni-collective-v4-frontier

Hugging Face datasets

Public dataset repos already published from this workspace:

  • Kai9987kai/supermix-conversation-datasets
  • Kai9987kai/supermix-science-vision-dataset

Benchmarks

The current local multibench comparison bundle is:

The current graph covers 20 benchmarked local model entries and keeps specialist-only models labeled separately when the common text suite is not the right evaluation fit.

Representative current common-benchmark leaders from the local graph JSON:

  • v33_final: 0.1867
  • v39_final: 0.1800
  • omni_collective_v1: 0.1633
  • v34_final: 0.1600
  • v36_native: 0.1533
  • v35_final: 0.1533
  • omni_collective_v4: 0.0900

Training entry points

Representative training and continuation scripts:

  • source/train_omni_collective_v2.py
  • source/train_omni_collective_v3.py
  • source/train_omni_collective_v4.py
  • source/train_omni_collective_v5.py
  • source/train_math_equation_model.py
  • source/train_image_recognition_model.py
  • source/build_reasoning_benchmix_v39.py
  • source/benchmark_all_models_common.py

If you want the active experimental path, start in source/.

Desktop build entry points

Primary desktop build helpers:

  • source/build_supermix_studio_desktop_exe.ps1
  • source/build_supermix_studio_desktop_installer.ps1
  • SupermixStudioDesktop.spec
  • installer/SupermixStudioDesktop.iss

The current bundled-model manifest is:

Research and experiment notes

This repo is a living experiment workspace. It contains finished artifacts, release-ready packaging, and in-progress work at the same time.

That means you will see mixed generations such as:

  • v28
  • v30
  • v33
  • v34
  • v35
  • v36
  • v37
  • v38
  • v39
  • omni_collective_v1 through omni_collective_v5

That is expected.

Recommended starting points

If you want to:

  • run a packaged local system
    • use runtime_python/
  • work on the active multimodel app
    • use source/supermix_multimodel_web_app.py
    • use source/supermix_multimodel_desktop_app.py
  • work on training
    • start in source/
  • inspect the current benchmark outputs
    • use output/benchmark_local_all_models_multibench_20260329.*
  • build a Windows installer
    • use the PowerShell build scripts in source/ plus installer/

Platform notes

  • Windows is the main desktop packaging target
  • the repo includes PyInstaller specs, PowerShell build scripts, and Inno Setup definitions
  • some training flows were designed around cloud GPU workflows, but the repo also supports local CPU experimentation

Security note

Do not commit or publish browser-session dumps, cookies, temporary automation state, or live access tokens.

Relevant policy docs:

  • SECURITY.md
  • CONTRIBUTING.md
  • CODE_OF_CONDUCT.md

License

See LICENSE.

Supermix

Supermix is the active monorepo for the Supermix desktop app, the Omni Collective training line, specialist model experiments, benchmark tooling, and local-first packaging.

It is intentionally a mixed workspace: source code, training scripts, build helpers, generated manifests, research outputs, and packaging metadata all live in the same repository so the full desktop + training workflow can be developed, tested, packaged, and published from one place.

Latest release

Supermix Studio X — V48 Frontier Edition is the latest desktop release.

The large installer and model bundles are hosted on Hugging Face due to repository asset size constraints and Git LFS quotas. The desktop release page links to the installer, model directory, and release assets.

What this repository contains

  • source/
    Active development workspace: model definitions, training scripts, dataset builders, benchmark runners, desktop/web app code, and release helpers.

  • runtime_python/
    A lighter packaged runtime path for quick local use.

  • datasets/
    Local conversation, coding, reasoning, science, and specialist training inputs.

  • output/
    Generated graphs, summaries, manifests, previews, logs, and Hugging Face upload staging.

  • installer/
    Inno Setup definitions and post-install notes for the desktop app.

  • dist/
    Locally built EXEs and installer outputs.

  • web_static/
    Lightweight browser-only metadata/static bundle.

Main capabilities

  • Multimodel desktop app with model selector, Auto routing, collective mode, and agent mode.
  • Local chat, omni, image, vision, math, protein, 3D, and materials specialist model families.
  • Curated desktop installer with built-in models plus one-click downloadable optional models.
  • Benchmark sweeps and local graph generation for the model zoo.
  • Hugging Face publishing helpers for models, installers, datasets, and downloadable model-store artifacts.
  • Training pipelines for omni, frontier, native-image, and specialist lines.

Current status

  • The latest finished full omni checkpoint in this workspace is omni_collective_v7.
  • omni_collective_v8 remains the long-running local experiment.
  • The desktop line uses a curated built-in bundle plus an in-app Hugging Face model store instead of shipping the full legacy model zoo.
  • The most recent desktop release is the V48 Frontier Edition.

Quick start

Run the packaged runtime

python runtime_python/chat_web_app.py

Windows launchers:

runtime_python\launch_chat_web_supermix.bat
runtime_python\launch_chat_terminal_supermix.bat

Run the active source web app

python source/chat_web_app.py

Run the multimodel desktop app from source

python source/supermix_multimodel_desktop_app.py

Run the training monitor

python source/training_monitor_gui.py --root .

Open the browser-only static bundle

Open:

web_static/index.html

Desktop downloads

Latest release assets

  • Release page: https://github.com/kai9987kai/Supermix/releases
  • Hugging Face desktop installer mirror: https://huggingface.co/datasets/Kai9987kai/supermix-studio-desktop-installer
  • Model zoo: https://huggingface.co/datasets/Kai9987kai/supermix-model-zoo

Local build outputs

  • dist/SupermixStudioDesktop/SupermixStudioDesktop.exe
  • dist/installer/SupermixStudioDesktopSetup.exe
  • dist/installer/SupermixStudioDesktopReleaseSHA256.txt

Curated bundle and model store

The desktop app ships with a curated core bundle and can also download optional models from the in-app model store.

The current workspace includes code and/or packaged artifacts for:

Qwen adapter line

  • v28
  • v30

Champion / frontier line

  • v31
  • v32
  • v33
  • v34
  • v35
  • v39
  • v40_benchmax
  • v41
  • v46
  • v47
  • v48

Native-image line

  • v36
  • v37
  • v38

Omni-collective line

  • v1
  • v2
  • v3
  • v4
  • v5
  • v6
  • v7
  • v8_preview
  • v8 in progress locally

Specialist lines

  • math_equation_micro_v1
  • science_vision_micro_v1
  • protein_folding_micro_v1
  • mattergen_micro_v1
  • three_d_generation_micro_v1
  • dcgan_mnist_model
  • dcgan_v2_in_progress

Latest finished omni model

The latest finished full omni checkpoint in this workspace is omni_collective_v7.

The latest preview checkpoint is omni_collective_v8_preview.

omni_collective_v8 is still training locally and is not yet the published stable omni release.

Benchmarks

Representative current results from the local benchmark inventory:

  • v40_benchmax: 0.2433 common benchmark score
  • omni_collective_v8_preview: 0.2167 on the reduced preview benchmark sweep
  • omni_collective_v7: 0.1067 on the current 6-benchmark local sweep

Specialist-only models remain in the graph inventory even when the common text benchmark is not the right evaluation fit.

Training entry points

Representative training and continuation scripts:

source/train_omni_collective_v2.py
source/train_omni_collective_v3.py
source/train_omni_collective_v4.py
source/train_omni_collective_v5.py
source/train_omni_collective_v6.py
source/train_omni_collective_v7.py
source/train_omni_collective_v8.py
source/train_math_equation_model.py
source/train_image_recognition_model.py
source/train_protein_folding_model.py
source/train_three_d_generation_model.py
source/train_mattergen_generation_model.py
source/benchmark_all_models_common.py

If you want the active experimental path, start in source/.

Desktop build entry points

Primary desktop build helpers:

source/build_supermix_studio_desktop_exe.ps1
source/build_supermix_studio_desktop_installer.ps1
SupermixStudioDesktop.spec
installer/SupermixStudioDesktop.iss

The current desktop app supports:

  • curated built-in bundle seeding
  • in-app Hugging Face model-store downloads
  • richer chat UI with drafts, context bank, compare bench, and dispatch preview
  • richer training monitor with recovery posture, rescue guidance, and fleet spotlight summaries

Hugging Face publishing

Public Hugging Face repos used by this workspace include:

  • Kai9987kai/supermix-model-zoo
  • Kai9987kai/supermix-studio-desktop-installer

Previously published model repos from this workspace include:

  • Kai9987kai/supermix-v33-frontier
  • Kai9987kai/supermix-omni-collective-v1
  • Kai9987kai/supermix-v38-native-image-xlite-fp16
  • Kai9987kai/supermix-v39-frontier-reasoning-plus
  • Kai9987kai/supermix-omni-collective-v2-frontier
  • Kai9987kai/supermix-math-equation-micro-v1
  • Kai9987kai/supermix-omni-collective-v4-frontier
  • Kai9987kai/supermix-omni-collective-v7-frontier

Public dataset repos include:

  • Kai9987kai/supermix-conversation-datasets
  • Kai9987kai/supermix-science-vision-dataset

Notes

  • This repo is intentionally not a clean source-only model repo.
  • Large local artifacts, checkpoints, and logs are often kept locally and mirrored selectively to releases or Hugging Face.
  • The safest way to consume the desktop app is through the curated-core release plus the in-app model store.

License

MIT



::contentReference[oaicite:1]{index=1}

Clone this wiki locally