#CoreML
TensorflowやらCoreMLやら触ることが増えたけどモバイルでできるのはパフォーマンス関連が主よなあ
February 27, 2024 at 4:11 PM
I am pitting 3 different ML implementations against each other (CoreML, CUDA, and Coral) to see which I like better.
November 14, 2024 at 6:52 AM
📦 ultralytics / ultralytics
⭐ 14,878 (+55)
🗒 Python

NEW - YOLOv8 🚀 in PyTorch > ONNX > OpenVINO > CoreML > TFLite
GitHub - ultralytics/ultralytics: NEW - YOLOv8 🚀 in PyTorch > ONNX > OpenVINO > CoreML > TFLite
NEW - YOLOv8 🚀 in PyTorch > ONNX > OpenVINO > CoreML > TFLite - GitHub - ultralytics/ultralytics: NEW - YOLOv8 🚀 in PyTorch > ONNX > OpenVINO > CoreML > TFLite
github.com
November 7, 2023 at 1:50 PM
今日のGitHubトレンド

argmaxinc/WhisperKit
このリポジトリは、OpenAIのWhisperモデルとAppleのCoreMLを組み合わせたSwiftパッケージ、WhisperKitを提供しています。
Swift Package Managerを使用してプロジェクトに追加できます。
詳細はREADMEとTestFlightのデモアプリ、ブログ記事を参照してください。
GitHub - argmaxinc/WhisperKit: On-device Speech Recognition for Apple Silicon
On-device Speech Recognition for Apple Silicon. Contribute to argmaxinc/WhisperKit development by creating an account on GitHub.
github.com
October 15, 2024 at 11:14 AM
I want to build an identifier app that works locally for better accuracy, speed, and with no connection required.
Over the next few days, I’ll be working with CreateML and CoreML. I’ve found a good dataset, and now it’s time to train the machine learning model.
#BuildInPublic
July 14, 2025 at 1:38 PM
You can trade these cards IRL to create a full collection of your friends.

I applied client-side AI models using MobileCLIP, CoreML, and on-device image classification to generate personalised stats from your own photos.
June 23, 2025 at 5:30 PM
And apparently Scrypted does this all with much less effort and uses CoreML on my mac mini *table flip*
November 13, 2024 at 7:33 AM
animagine xl 3.0のCoreML変換モデル
Stable Diffusion CoreML ml-stable-diffusion
Mochi Diffusion v5.0
February 15, 2024 at 3:27 PM
Some TextKit fun - made a recognizer to find questions/answers in arbitrary text (CoreML model with some regex overrides), and UITextView applies style to the question and answers (real time view of the output cards on the right). Should make it easier to import a bunch of cards. #iosdev #BenkyoBox
November 28, 2024 at 6:50 AM
If you're interested in end to end creating and training a model for CoreML, this tutorial by @kingreza is great. http://www.reza.codes/2017-07-29/how-to-train-your-own-dataset-for-coreml/
November 23, 2024 at 12:31 PM
Today I'm benchmarking CoreML models 🤓

First, iOS simulation; then, AWS Device Farm.

❓ Does anyone know of a device farm for PCs?

Pd: the model is a ~80 MB FastVIT, pretrained by Apple for classification. It's surprising how fast it runs!
January 16, 2025 at 11:41 AM
This week, I've finished the "Better Rest" app of #100DaysOfSwiftUI. A very simple app just to learn about some concepts of State, DatePicker, Stepper and CoreML. So, I've decided to explore about #AppleWatch using this app, and I want to make some questions to the experts. I'll start...
July 31, 2024 at 10:54 PM
Want to learn how to develop an image classification app with swift and CoreML?

In the third post in the series we finish a first version of the ZooScan UI and provide a basis for adding a ML classifier later.

thinkpractice.nl/post/zooscan...

#swift #swiftlang #iOSDev #Buildinpublic #ML #coreml
ZooScan – Part 3: Storing Our Scanned Animals and Finalizing the UI
In this third part of the ZooScan series, we implement the ViewModel and ScannedAnimal model, and create the Main and Detail views. This sets the stage for the machine learning classification in futur...
thinkpractice.nl
June 14, 2025 at 6:40 AM
#SituationalAwareness: every week is a massive week.

x.com/reach_vb/sta...
November 24, 2024 at 10:27 PM
것은 연말에 나올 맥 시리즈일 것으로 봄.

말 그대로 물들어 오니 열심히 노 젓는 애플을 보고 있는 건데, 그래도 m4가 이렇게 빨리 나오리라곤 예상치 못했다. 소프트웨어 지원까지 보면 생각보다 관련한 준비 기간은 길었던 거 같음. coreML 지원은 이미 2년 전부터 공개 리포지토리에서 제법 보인 만큼 티가 아예 안 난 건 아니지만서도.

펜슬 프로의 경우 펜슬 2에 있던 터치 기능을 제대로 써먹게 만들려고 햅틱 넣어둠. 사실 원래부터 이렇게 했어야 한다 보는 게, 펜슬 2 터치 기능은 진짜 오동작이 너무 심해서 난 사실상
May 7, 2024 at 11:19 PM
发现了一个可以在iPhone上画AIart的App:Draw Things
体积小,支持导入模型,随时随地作画
不过用起来很耗电(CoreML),模型需要转换成checkpoint文件格式,而且操作起来比起WebUI稍微麻烦一些。
#furry #furryai #furryart #AIart
December 11, 2024 at 3:39 AM
this is actually an important question since so many devices people are running games on today have silicon whose API is ONNX (or something you can compile to from ONNX like apple coreml)
January 16, 2025 at 5:21 PM
(reddit) Microsoft's deep learning neural network model on Apple's CoreML for iOS, recognizing visual objects https://www.reddit.com/r/videos/comments/6iv4wg/googles_new_vision_api_is_unbelievably_fast/
Reddit - The heart of the internet
www.reddit.com
April 11, 2025 at 2:34 AM
Linear regression and recommender algorithms on-device in CoreML? Consider me excited.
November 19, 2024 at 10:12 AM
Included in this list are the MobileCLIP models, Apple’s own implementation of OpenAI’s CLIP neural network, optimised for Core ML and iOS.

Learn to implement client-side AI with CoreML and CLIP models in the full article: blog.jacobstechtavern.com/p/offline-ai...
Customised client-side AI with Apple's CLIP models
Perform Core ML magic, entirely offline
blog.jacobstechtavern.com
June 11, 2025 at 12:30 PM
Threads/IG did (?does) something very similar. As of the last time I conducted a forensic analysis of Threads (several months ago), they shipped the iOS app with at least 2 models and leveraged CoreML to personalize feeds on-device. The challenge becomes cross-platform feed state synchronization
June 12, 2025 at 5:15 AM
自分もM1 MacでCoreML対応版に差し替えてlargeモデルを食わせようとしたらCore MLモデルの生成が全然終わらなかったでござる
whisper.cppのCore ML版をM1 MacBook Proで動かす
zenn.dev
May 5, 2023 at 3:05 AM
생각해보니 whisper작동이 오래걸리는게 맥북에어 램이 8기가라 램때문에 그런가..

일단 램때문에 그런거 같으니, 라지모델은 보류하고 일단 스몰 모델로 진행. core ML 컴파일이 2시간정도 걸릴 것 같으니 일단 오늘 중에는 안될꺼고.. 자기 전에 시도해볼 수 있을라나 모르겠다.

일단 coreML 에서 스몰 모델로 작업시 매번 초기화를 한다 하면, 라지모델은 현재 맥북에서 사용할 수 없는걸로.
아 이러면 환불해야하나...
September 8, 2023 at 2:22 PM
5️⃣ Compiling on-device
Once downloaded, the CoreML model is optimized on the user’s device, ensuring peak performance for their hardware. Translation? Faster, smoother apps.
December 3, 2024 at 12:08 PM