#风景[超话]#
不用晕船！一篇文章告诉你9-10月蕞好观鲸点！
Hello宝子们，悉尼的宝子是幸运的哈！我们不用特别远行就能观赏到鲸鱼迁徙。

9-10月是蕞后的时间~天气也很暖和~！大家都出去走走‍♀️吧！（部分地区之前发过游玩攻略的，合集）

——
Hornby Lighthouse
Watsons Bay

-
Ben Buckler Point
Bondi Beach

-
Bare Island
La Perouse

-
Burragala Lookout
North Head

-
The Gap
Watsons Bay

-
The Bondi to Coogee Walk
Bondi-Coogee

-
Barrenjoey Headland
Palm Beach

-
Long Reef Point
Collaroy

-
The Coast Track
Royal National Park

-
Mala疤r Headland National Park
Anzac Parade, Mala疤r NSW 2036

这些地方都超美的，即便是去徒步和欣赏美景都值得！
#沿途的风景##风景#

AIGC text to image

prompt = "A collage of four images, each showing a different type of storm. The first image shows a satellite view of \
a hurricane over the ocean, with swirling clouds and a visible eye. The second image shows a tornado touching down on a \
rural landscape, with debris flying in the air. The third image shows a blizzard covering a city in snow, with strong \
winds and low visibility. The fourth image shows a thunderstorm over a mountain range, \
with lightning bolts and dark clouds."
image = pipe(prompt, num_inference_steps=4, guidance_scale=8.0).images[0]
image

-------------
sdxl_lcm_diffusers_colab.ipynb
ttps://colab.research.google.com/github/camenduru/latent-consistency-model-colab/blob/main/sdxl_lcm_diffusers_colab.ipynb

!pip install -q diffusers transformers accelerate peft

from diffusers import UNet2DConditionModel, DiffusionPipeline, LCMScheduler
import torch

unet = UNet2DConditionModel.from_pretrained("latent-consistency/lcm-sdxl", torch_dtype=torch.float16, variant="fp16")
pipe = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", unet=unet, torch_dtype=torch.float16, variant="fp16")

pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
pipe.to("cuda")

新设备：表面活性化接合装置；制造商：日本制造商
实现了在常温下能得到块状破坏强度的晶圆接合。从研究开发用接合装置到量产用接合装置都有对应。在晶圆级封装、三维安装等广泛领域使用。
SAB-60R简介：SAB-60R是一种在真空中输送晶片，对准后可接合的批量生产用表面活性化接合装置。采用盒式磁带to盒式磁带方式，接合后晶片被搬送到搬出用盒式磁带上。对准可以选择使用外形对准和标记的方式。