Skip to content
Other

MagicEightBall FP32 Base

a compact model for policy checking and text classification

2 min read

The MagicEightBall model evaluates a fixed support policy and returns a boolean response, a category, and a suggested action. The base FP32 file is only 1,832 bytes, and the included CPU runner supports processing up to 131,072 Needle 3 tokens. The graph reads seven FP32 facts and returns seven corresponding logical inferences. This tool is useful for developers who need ultra-lightweight text prompt checking and rule-based routing without heavy models.

Key features

  • Full precision: a dense logical reference model based on FP32 that compiles support rules.
  • Large context length: supports incoming sequences of up to 131,072 Needle 3 tokens.
  • Ready-to-use CPU runner: includes a CLI for quick validation and local execution.
  • Proven reliability: tested across 128 fact assignments, 128 short text fixtures, and spot documents.

What you can do

  • Run in a local CLI: execute the script via Python for rapid request classification (for example, in billing).
  • Check text rules: evaluate whether a prompt falls under specific policies using the built-in engine.
  • Test ONNX integration: use the model as an ultra-compact example for ONNX Runtime.

Example

python -m pip install numpy sentencepiece onnxruntime
python portable_runtime.py model.onnx --text "Duplicate charge. Review authorized."
Read moreDetails
Benchmarks: passed 128 facts and 128 text fixtures, deviation above 131,073 tokens.
Hardware requirements: extremely low (the model weighs 1,832 bytes, a CPU runner is used).
License: decision rules and runner are distributed under the MIT license, Cactus Compute Needle 3 tokenizer under Apache-2.0.
Examples
Discuss Telegram 0

Will it run?

Runs in BF16

Fits without quantisation.

  • BF16weights 0.0 GB · ready files0.5 GB
  • FP8 / INT8recommendedweights 0.0 GB0.5 GB
  • 4-bitweights 0.0 GB0.5 GB

Tip: go 8-bit.FP8 / INT8 is within about a percent of full precision in quality, takes half the memory and runs faster on RTX 40/50 and H100. Full precision only pays off for fine-tuning.

What to get for this model

Massed ComputeOpen Massed Compute

Your own GPU: FP8 / INT8 — from RTX 2060 · BF16 — from RTX 2060

Massed Compute hourly prices, November 2026. The discounted price applies with the coupon at checkout.

Estimate: weights plus working memory. Long videos, high resolutions and long contexts need more.

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # chadstthomas/MagicEightBall-FP32
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download chadstthomas/MagicEightBall-FP32 --local-dir ~/models/MagicEightBall-FP32
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to MagicEightBall-FP32
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

Model family

New posts about versions of MagicEightBall-FP32 (quants, finetunes, LoRAs) will show up under the bell.

No known relatives of this model yet. All derivatives on Hugging Face

About the model

chadstthomas/MagicEightBall-FP32
Author
chadstthomas
Category
Other
Released
Posted

On NeyroLearn

Hugging Face

00

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments