Skip to content
Other

orchid-clip-v9cw05

CLIP-энкодер для определения видов орхидей

1 min read

orchid-clip-v9cw05 от musharna — image encoder для фотографий орхидей, дообученный из BioCLIP 2.5 (ViT-H/14). Модель выдаёт 1024-мерные L2-нормализованные признаки изображения. Это второй из двух энкодеров в Space orchid-genus-id, где его оценки усредняются с orchid-clip-v8.

Ключевые особенности

  • Кодирует фото в 1024-мерный L2-нормализованный вектор для сравнения по косинусу.
  • Основа — BioCLIP 2.5 ViT-H/14; дообучение на 1,12 млн изображений 5 124 видов орхидей.
  • Веса — среднее fine-tuned и base модели 50/50 по WiSE-FT, отсюда «w05».
  • В Space усредняется с orchid-clip-v8: species top-1 0.742, genus top-1 0.937.
  • С abstain-правилом при 90% precision называет вид в 66% случаев.
  • Чекпойнт весит 3,9 ГБ, CPU-инференс медленный.

Пример

pip install open_clip_torch torch pillow
hf download musharna/orchid-clip-v9cw05 --local-dir orchid-clip-v9cw05
python orchid-clip-v9cw05/embed_example.py orchid-clip-v9cw05 your_orchid.jpg

Важно: загружайте с pretrained=None и затем state dict; имя hf-hub: подтянет base BioCLIP 2.5, а не эти веса.

ПодробнееНа тесте 8 837 фото 3 369 видов: species top-1 0.723, genus top-1 0.935, вид при 90% precision — 62%; в паре с v8 — 0.742 / 0.937 / 66%.
Оценён только image tower; text tower для zero-shot не тестировали.
Ошибки чаще внутри рода; гибриды не целевые.
Код и веса — MIT, но ~70% обучающих фото — CC BY-NC: учитывайте для коммерческого использования.
Discuss Telegram 0 40

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.

    # musharna/orchid-clip-v9cw05
    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 musharna/orchid-clip-v9cw05 --local-dir ~/models/orchid-clip-v9cw05
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to orchid-clip-v9cw05
  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 orchid-clip-v9cw05 (quants, finetunes, LoRAs) will show up under the bell.

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

About the model

musharna/orchid-clip-v9cw05
Author
musharna
Category
Other
Released
Posted

On NeyroLearn

Hugging Face

400

Ask AI about this model

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

Comments