接下來想來試試用產圖，Gemini 現在可以畫圖了(用中文指令會沒反應，得用英文)。

我想著 API 肯定也可以用了吧，找了半天發現他只開放在 Vertex AI 上，得另外計價，而且要用還得要填表申請，猶豫了很久決定先不試，直接去用我已經買了 plus 的 ChatGPT。
https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/imagen-api
https://cloud.google.com/vertex-ai/generative-ai/docs/image/overview#feature-launch-stage
https://cloud.google.com/vertex-ai?hl=zh-TW#pricing

於是我開啟 API 頁面看，發現他特別提示了 API 帳單跟 ChatGPT 是分開的，我這才意識到他是 open ai API，本來就不一樣，想的太美了。

猶豫再三，決定還是先找個開源的模型試試，免費的測完再去玩付費的。想著去年試過的 stable diffusion 效果還不錯，乾脆再把他架起來試試，於是上網搜尋一下目前發展到甚麼地步。
https://platform.stability.ai/pricing

發現發展到要收費了...
不過還好舊版本的模型有開源，從開源版本中選了效果不錯的 Stable Cascade 來試試。
Stable Cascade 有發佈在 huggingface 上，那要測就很單純了，甚至連範例程式都寫好了。
https://huggingface.co/stabilityai/stable-cascade

這次一樣到 colab 架，選一個有 GPU 的環境，我跟之前一樣選了 T4 GPU。

照著做就行了，先把 huggingface 的 library，diffusers 裝起來。

pip install diffusers

Collecting diffusers
Downloading diffusers-0.30.0-py3-none-any.whl.metadata (18 kB)
Requirement already satisfied: importlib-metadata in /usr/local/lib/python3.10/dist-packages (from diffusers) (8.2.0)
Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from diffusers) (3.15.4)
Requirement already satisfied: huggingface-hub>=0.23.2 in /usr/local/lib/python3.10/dist-packages (from diffusers) (0.23.5)
Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (from diffusers) (1.26.4)
Requirement already satisfied: regex!=2019.12.17 in /usr/local/lib/python3.10/dist-packages (from diffusers) (2024.5.15)
Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from diffusers) (2.32.3)
Requirement already satisfied: safetensors>=0.3.1 in /usr/local/lib/python3.10/dist-packages (from diffusers) (0.4.4)
Requirement already satisfied: Pillow in /usr/local/lib/python3.10/dist-packages (from diffusers) (9.4.0)
Requirement already satisfied: fsspec>=2023.5.0 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub>=0.23.2->diffusers) (2024.6.1)
Requirement already satisfied: packaging>=20.9 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub>=0.23.2->diffusers) (24.1)
Requirement already satisfied: pyyaml>=5.1 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub>=0.23.2->diffusers) (6.0.2)
Requirement already satisfied: tqdm>=4.42.1 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub>=0.23.2->diffusers) (4.66.5)
Requirement already satisfied: typing-extensions>=3.7.4.3 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub>=0.23.2->diffusers) (4.12.2)
Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.10/dist-packages (from importlib-metadata->diffusers) (3.20.0)
Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests->diffusers) (3.3.2)
Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->diffusers) (3.7)
Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests->diffusers) (2.0.7)
Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->diffusers) (2024.7.4)
Downloading diffusers-0.30.0-py3-none-any.whl (2.6 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 2.6/2.6 MB 29.6 MB/s eta 0:00:00
Installing collected packages: diffusers
Successfully installed diffusers-0.30.0

之前測試時有時候會發現環境沒有切換正確，所以我會先確認一下 torch 版本，還有 GPU 是不是能被使用的。不然等下載模型完才發現，又得刷新環境重來，會浪費時間。

import torch
from diffusers import StableCascadeDecoderPipeline, StableCascadePriorPipeline
print(torch.__version__)
print(torch.cuda.is_available())

2.3.1+cu121
True

把模型下載下來。

prior = StableCascadePriorPipeline.from_pretrained("stabilityai/stable-cascade-prior", variant="bf16", torch_dtype=torch.bfloat16)
decoder = StableCascadeDecoderPipeline.from_pretrained("stabilityai/stable-cascade", variant="bf16", torch_dtype=torch.float16)

給 prompt 讓模型生成影像。

prompt = "an image of a shiba inu, donning a spacesuit and helmet"
negative_prompt = ""
prior.enable_model_cpu_offload()
prior_output = prior(
    prompt=prompt,
    height=1024,
    width=1024,
    negative_prompt=negative_prompt,
    guidance_scale=4.0,
    num_images_per_prompt=1,
    num_inference_steps=20
)
decoder.enable_model_cpu_offload()
decoder_output = decoder(
    image_embeddings=prior_output.image_embeddings.to(torch.float16),
    prompt=prompt,
    negative_prompt=negative_prompt,
    guidance_scale=0.0,
    output_type="pil",
    num_inference_steps=10
).images[0]
decoder_output.save("cascade.png")

可以把圖打開看看效果，還算順利。

換個 prompt 試試，效果不錯。

prompt = "cute black cat, Wearing a wizard hat. Cartoon."