Skip to content
Other

RMSN-243M

компактная SFT-LLM для CoT-рассуждений

1 min readupdated

RMSN-243M-Bulk-Boundary-SFT — 243,73M-модель от DuoNeural на кастомном Port-Hamiltonian-стеке. Это SFT-базовая линия Phase 5, дообученная на reasoning-трассах DuoNeural (CoT SFT, 15k шагов). Архитектура: hidden dim 1024, 16 слоёв, 64 workspace slots, токенизатор cosmo2 с vocab 49 152.

Ключевые особенности

  • Архитектура Resonant Metriplectic Sheaf Network с Port-Hamiltonian-формулировкой.
  • Солитон мигрирует из bulk L8 в boundary L15 по мере SFT; на 15k шагов фиксируется на L15.
  • Обучение на CoT-трассах DuoNeural; cot_ppl 10,31, wiki_ppl 184,81.
  • Удержание 2,6× относительно базы после 15k шагов SFT.
  • Топология pointlike soliton, по карточке, снижает поверхность атаки dual-core tube-топологии.

Пример

Сценарий из карточки: генерация цепочек рассуждений после CoT SFT на reasoning-трассах DuoNeural. Для этого режима указана ключевая метрика cot_ppl = 10,31 при 15k шагов SFT.

ПодробнееБенчмарки SFT (15k шагов): wiki_ppl 184,81; cot_ppl 10,31; retention ratio 2,6× base.
Параметры: 243,73M, 16 слоёв, hidden dim 1024, 64 workspace slots, vocab 49 152.
Обучение: CoT SFT на 15k шагов, DuoNeural Phase 5; язык — английский.
VRAM и ограничения в карточке не указаны.
Discuss Telegram 0 19

Will it run?

Runs in BF16

Fits without quantisation.

  • BF16weights 0.7 GB · ready files1.3 GB
  • FP8 / INT8recommendedweights 0.4 GB0.9 GB
  • 4-bitweights 0.2 GB0.7 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.

    # DuoNeural/RMSN-243M-Bulk-Boundary-SFT
    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 DuoNeural/RMSN-243M-Bulk-Boundary-SFT --local-dir ~/models/RMSN-243M-Bulk-Boundary-SFT
    # The model is in ~/models. How to run it is in the model card (link above).
  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 RMSN-243M-Bulk-Boundary-SFT (quants, finetunes, LoRAs) will show up under the bell.

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

About the model

DuoNeural/RMSN-243M-Bulk-Boundary-SFT
Author
DuoNeural
Category
Other
Released
Posted

On NeyroLearn

Hugging Face

190

Ask AI about this model

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

Comments

Other

BowPose

BowPose