Skip to content
Other

Sage-SciPerception

научные данные на входе LLM

1 min readupdated

Sage-SciPerception от InternLM — набор обученных projectors, которые отображают научные представления в пространство эмбеддингов языковой модели. Поддерживаются кристаллы, спектры, биологические последовательности и биосигналы; опубликованные чекпоинты собраны для glm-5.3-flash.

Ключевые особенности

  • Соединяет научные модальности с LLM через projectors; научные модели и LLM при обучении остаются замороженными.
  • Задачи: свойства материалов, разметка PPI-сайтов, идентификация молекул по спектрам, классификация ЭКГ/ЭЭГ, splice-site и семейства РНК.
  • Готовые веса для CIF, protein-seq, spectra, ECG, DNA, EEG, RNA, XRD и protein-struct.
  • Каждый чекпоинт хранит совместимую языковую модель и формат входа в config.json.
  • Инференс через vLLM с prompt_embeds: проекция научных признаков выполняется на клиенте, генерация — на сервере.

Пример

Проверка загрузки projector’а на CPU из карточки:

from load_projector import load_projector

projector, config = load_projector(
    "projectors/glm-5.3-flash/cif/jarvis", device="cpu"
)
print(config["task"])
print(sum(p.numel() for p in projector.parameters()))
print(config["output_dim"])
ПодробнееБенчмарки: JARVIS-DFT formation energy — MAD/MAE 15.03 (больше — лучше); MPBind PPI — per-protein F1 61.12%; BEACON ncRNA — accuracy 97.35%.
Для inference нужен vLLM с --enable-prompt-embeds; в примере — tensor-parallel-size 8 и max-model-len 8192. VRAM не указан.
Ограничения: лицензия не указана; task-adapted научные модели для DNA, MPBind и ncRNA не входят; пример стартует с уже подготовленных scientific features.
Энкодеры и preprocessing — отдельные зависимости.
Видео-пример
Discuss Telegram 8 59

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.

    # internlm/Sage-SciPerception
    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 internlm/Sage-SciPerception --local-dir ~/models/Sage-SciPerception
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to Sage-SciPerception
  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 Sage-SciPerception (quants, finetunes, LoRAs) will show up under the bell.

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

About the model

internlm/Sage-SciPerception
Author
internlm
Category
Other
Released
Posted

On NeyroLearn

Hugging Face

598

Ask AI about this model

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

Comments

Other

NeCast

NeCast