В прошлом посте говоря "Все вызовы теперь одинаковы" я несколько слукавил. Всё-таки есть в этом зоопарке версий некоторая несовместимость вызов которой просто так не унифицировать. Эти моменты вынесены в отдельный модуль QtCompat (compatibility). Там не так много функций но они довольно полезны.
Этот модуль содержит унификаци модуля shiboken2, функций loadUi, translate и несколько переименованных функций классов или изменённую сигнатуру аргументов и возвращаемых значений. Это единственное исключение из правила когда вам потребуется где-то изменить свой код кроме импортов и этот код не похож на обычный код PySide2.
Например, в PyQt4 и PySide есть метод
QHeaderView.setResizeMode
Для PyQt5 и PySide2 они были благополучно переименованы в
QHeaderView.setSectionResizeMode
Чтобы применить этот метод следует использовать такой код
from Qt import QtCompath
header = self.horizontalHeader()
QtCompat.QHeaderView.setSectionResizeMode(header, QtWidgets.QHeaderView.Fixed)
Унификация загрузки UI файлов:
# PySide2
from PySide2.QtUiTools import QUiLoader
loader = QUiLoader()
widget = loader.load(ui_file)
# PyQt5
from PyQt5 import uic
widget = uic.loadUi(ui_file)
# Qt.py
from Qt import QtCompat
widget = QtCompat.loadUi(ui_file)
Хорошо что таких моментов не много и их легко запомнить.
Полный список можно посмотреть в таблице.
#qt#tricks
#ainews
Google labs "Pomelli" productini experiment qilib ommaga taqdim qildi
Oldin juda cheklangan odamlar, cheklangan geografik hududda ishlatsa bo'lardi, hozir 170 ta davlatga imkon ochildi
Xo'sh, bu qanday dastur:
Bu googlening "AI marketing" instrumenti
U sizga kompaniyangizni DNA'ni aniqlab beradi, qanday turdagi kompaniyasiz, nima elementlaringiz bor, qanday ranglar sizni ranglaringiz...
Va asosiy joyi, sizga o'zingizga moslab boshqa kampaniya va o'zgarishlar taqdim qiladi
Xullas, sinab ko'ringlar, tushunib olasizlar
Note: U hozir sinov tariqasida ishlayotgani uchun mutlaqo tekin🙃
Link: https://labs.google.com/pomelli
Boshqa ko'plab yangiliklar uchun obuna bo'ling
@hikoyachii
#AInews
Midjourney released their latest v6.1 model, an improvement over the pervious model v6.0🌴
🚩Coherence and Detail: Better representation of human anatomy and small features like eyes and hands.
🚩Image Quality: Reduced pixel artifacts and enhanced textures, particularly for skin and retro styles.
3️⃣Speed Improvements: Standard jobs are approximately 25% faster.
4️⃣Text Accuracy: Improved rendering of text within images when using quotations in prompts.
5️⃣Personalization Model: Enhanced nuance and accuracy, allowing users to apply previous personalization codes.
6️⃣New Upscalers: Introduces advanced upscalers for better image texture quality.
7️⃣New Mode: The --q 2 mode adds texture but may reduce coherence.
@ai_artz🆒
Boost⚡️
#AInews
LivePortrait: Portrait Animation Generator
LivePortrait is an online platform that allows users to turn static images or photographs into animated videos using AI technology. The platform uses specially trained models to transfer movements from video or webcam footage onto images, enabling users to animate portraits and headshots.
The key features of LivePortrait include it Customization and its training data of around 69 million high-resolution video frames as training data, along with advanced keypoint technology to deliver realistic facial animations.
It is eas and free to use to quickly turn their existing photos and images into animated videos in just seconds.
It is worth trying given its ups and downs. But as an example I tried it for myself with Bella Poarch and Adolf Hitler 😁. You can access Live Portrait inside Huggingface.
🤖Access LivePortrait
📚Read more
🌿Join us @Ai_artz for more cool AI tools.
#AInews
Odysseyml; Hollywood-grade Visual AI
Odyssey, a startup around a year old, is building Hollywood-grade visual AI that can generate cinematic scenery, characters, and lighting with fine-tuned control over every element in a scene.
Unlike current text-to-video tools that produce fully-baked video files, Odyssey has trained four generative models that allow users to have precise control over the geometry, materials, lighting, and motion in their videos.
The founders argue that for AI to work for Hollywood, it needs to create "glitch-free and mind-blowing visuals". To achieve this, Odyssey has secured $9 million in funding and is leveraging technologies developed for autonomous vehicles, which required highly accurate visual AI and physics simulations.
🌿@Ai_artz
📚Read More
#AInews
Here's something to takecare if you spend some of your time on the internet. We all know that with the rise of text-to-image technology, AI has sparked significant debate among the public.
While it has brought substantial benefits by simplifying certain work processes, its side effects have left some of us feeling skeptical.
The video above demonstrates how people are using AI to create deepfake videos for deceptive ads. It features Joe Rogan, Mr. Beast, Andrew Tate, and Andrew Huberman.
Watch the full video on YouTube 📺
https://youtu.be/-e6q3lPqjBo?si=aivp0egSKmxevHMn
👓Read More in our Telegram channel; Click
🌿--- --- --- --- --- --- --- --- --- --- --- --- -- ---🍒
🇪🇸The Shocking AI Images in Spain Highschool
⚡️How Adobe is Upgrading Itself Using AI
🤖 A Powerful yet FREE AI Image Generator Website
🪟Microsoft is Integrating AI to Window 11 and Microsoft 365
🌿--- --- --- --- --- --- --- --- --- --- --- --- -- ---🍒
✅ Decoy Voice _YouTube
Join us here 👉@ai_artz📢
#AInews
Here is how Adobe is upgrading itself using text to image AI technology.
🎥 Watch full video on YouTube:
https://youtu.be/VY3cPeiskKI?si=qTUX7XTHi_ltar7t
✅ Matt Wolfe _YouTube
@ai_artz📢
Месяц с лишним назад я перегрузился новостями про AI и перестал читать про это. Надо вкатываться обратно. Тег будет #ainews, поехали. Один-два поста в неделю отныне посвящяются новостям по ИИ, которые кажутся важными мне. С моими комментариями.
❗️ Вы не представляете, НАСКОЛЬКО сложной оказалась задача сделать обзор событий хотя бы за последнюю неделю. Это заняло у меня 3 (три!) (ТРИ!) часа. Надеюсь результат достоин усилий и заслуживает лайка и репоста 👇
⚡️Anthropic представила Claude 4 Opus и Sonnet 4
На презентации Code/w Claude показали две новые модели:
Opus 4 — флагман для разработчиков: справляется с многошаговыми задачами, держит контекст часами (в демо — игра в Pokémon), понимает большие кодбазы.
Sonnet 4 — доступен даже в бесплатном плане. Улучшен по точности, устранил ошибки в код-навигации и стал быстрее.
Обе модели:
— умеют чередовать анализ с веб-поиском,
— работают параллельно,
— интегрированы в VS Code, JetBrains и GitHub.
Тарифы прежние:
Opus 4 — $15/$75 за млн токенов
Sonnet 4 — $3/$15
Я не особо пользуюсь, но если хочется что то кроме open ai, антропик очень хорош )
#ainews | Acidcrunch
Любой текст в подкаст! Новая бесплатная фича NotebookLM от Google
Теперь можно быстро делать аудио в стиле подскаста для любых текстов. Это поможет изучать новые предметы на ходу или просто озвучивать интересные статьи.
Тестить тут
👇 Подпишись, сохрани и кидай друзьям
#ainews | AcidCrunch
AI and Data Analytics news for December 2023
A recent forecast from IDC shows that the worldwide artificial intelligence software market is set to grow from $64 billion in 2022 to nearly $251 billion in 2027 at a compound annual growth rate of 31.4%. This means that we will continue to closely monitor AI news in monthly digests.
To round up the year of AI and data analytics developments, let’s see how the market has grown and if there are new projects that bring valuable results. #AInews
Read the article
#from#AInews
Microsoft reveals Copilot: “Everyday AI companion"
Microsoft just went HAM at their Microsoft Surface and AI event, announcing Microsoft 365 Copilot and 365 Chat, and significant updates to Windows 11 and Bing.
🌿--- --- --- --- --- --- --- --- --- --- --- --- -- ---🍒
Microsoft 365 Copilot/Chat:
Microsoft introduces new capabilities for Copilot in Outlook, Word, Excel, Loop, OneNote, and OneDrive.
Microsoft reveals 365 Chat — search across ALL your info, including emails, meetings, chats, documents, and more, plus the web.
Like an assistant, 365 Chat deeply understands you, your job, your priorities, and your organization and can assist with tedious tasks.
🌿--- --- --- --- --- --- --- --- --- --- --- --- -- ---🍒
Windows 11:
Microsoft unveils Copilot directly integrated into Windows 11 coming for free on Sept. 26th.
Windows 11 apps like Word, Paint, and Edge also gain AI features like background removal, image creation, and enhanced search.
🌿--- --- --- --- --- --- --- --- --- --- --- --- -- ---🍒
Bing:
Bing will make the latest OpenAI text-to-image model DALL-E 3 accessible all for free (demo).
Bing will also have more personalized answers based on history, a new AI shopping experience, and updates to Bing Chat Enterprise.
🌿--- --- --- --- --- --- --- --- --- --- --- --- -- ---🍒
Why it matters: We are entering an entirely new era with AI. Microsoft 365 Copilot is like having a personalized assistant helper who knows everything about you and your work, at any time.
✅ Source: The Rundown.ai
@ai_artz📢
#from#AInews
A sleepy town in southern Spain is in shock after it emerged that AI-generated naked images of young local girls had been circulating on social media without their knowledge.
The pictures were created using photos of the targeted girls fully clothed, many of them taken from their own social media accounts.
These were then processed by an application that generates an imagined image of the person without clothes on.
So far more than 20 girls, aged between 11 and 17, have come forward as victims of the app's use in or near Almendralejo, in the south-western province of Badajoz.
"One day my daughter came out of school and she said 'Mum there are photos circulating of me topless'," says María Blanco Rayo, the mother of a 14-year-old.
"I asked her if she had taken any photos of herself nude, and she said, 'No, Mum, these are fake photos of girls that are being created a lot right now and there are other girls in my class that this has happened to as well.'"
She says the parents of 28 girls affected have formed a support group in the town.
Police are now investigating and according to reports, at least 11 local boys have been identified as having involvement in either the creation of the images or their circulation via the WhatsApp and Telegram apps.
✅ BBC
Click here to continue reading all the articles
@ai_artz