Skip to content
Computer vision

Lucia v1

9B-адаптер, который проговаривает картинку по шагам

2 min read

Lucia v1 от AbstractPhil — старшая в семействе Lucia, LoRA-адаптер к Qwen/Qwen3.5-9B (image-text-to-text). Модель смотрит на изображение и выдаёт строку «thought-symbolic-image»: короткую цепочку мыслей, по одному именованному шагу на строку, затем подпись, prompt и теги, а после — вопросы, которые задал бы внимательный зритель, с ответами и пометкой, чем ответ подтверждён.

Ключевые особенности

  • Фиксированный набор шагов цепочки: see, name, place, count, relate, form, light, colour, mood, mean, intend, story, compare, change, check, say.
  • Глубина взгляда задаётся словами в промпте: «a quick look» (1–2 мысли), «a closer look» (3–5), «a long look» (6–10).
  • LoRA rank 32, alpha 64, dropout 0.05, 80.2M обучаемых параметров; vision tower заморожен.
  • Каждый вопрос помечается: [record] — факт с картинки, [chain] — выведено цепочкой, [unknown] — неизвестно.
  • Только английский; thinking выключен и в обучении, и в инференсе — нужны transformers 5 и peft 0.18+.

Пример

from PIL import Image
image = Image.open("picture.jpg").convert("RGB")
prompt = ("Look at this picture and think it through in writing, a closer look, ending with a caption; then write a caption and a prompt or tags, and the questions a careful looker would still ask.")
messages = [{"role": "user", "content": [{"type": "image", "image": image}, {"type": "text", "text": prompt}]}]
text = processor.apply_chat_template(messages, add_generation_prompt=True, tokenize=False, enable_thinking=False)
inputs = processor(text=[text], images=[image], return_tensors="pt").to(model.device)
out = model.generate(**inputs, max_new_tokens=800, do_sample=False)
Подробнее500 отложенных картин, медиум назван верно из 100: рисунок 86, объект 90, картина 83, фотография 80, эстамп 76 (у прошлой версии — 80 / 4 / 20 / 76 / 67).
200 музейных картин: 13 из 27 картин, 38 из 39 эстампов, 66 из 66 объектов.
0.27 с на строку на одной RTX PRO 6000, batch 64, 800 новых токенов.
Ограничения: быстрый взгляд хрупок, строки-спецификации без картинки дают определяющую мысль примерно в 1 случае из 10, язык только английский. Лицензия Apache-2.0; для 4B, 2B и 0.8B есть и полные файнтюны.
Discuss Telegram 0 111

Will it run?

Working out the memory needs

The bot is still collecting this model's weight sizes; the estimate will appear here in a few minutes. Meanwhile you can rent a GPU with headroom.

Massed ComputeOpen Massed Compute

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.

    # AbstractPhil/qwen3.5-9b-think-symbolic-image
    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 AbstractPhil/qwen3.5-9b-think-symbolic-image --local-dir ~/models/qwen3.5-9b-think-symbolic-image
    # 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 qwen3.5-9b-think-symbolic-image (quants, finetunes, LoRAs) will show up under the bell.

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

About the model

AbstractPhil/qwen3.5-9b-think-symbolic-image
Author
AbstractPhil
Category
Computer vision
Released
Posted

On NeyroLearn

Hugging Face

1110

Ask AI about this model

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

Comments