Tired of training models that never see the light of day? Don't let your hard work go to waste!

In this 🧵, I'll show you how to pick from over 900+ models from TIMM by @wightmanr , train them with Fastai by @jeremyphoward , and deploy them on Android – all for free. ⚡ You will learn how to:

▶ Load a model from TIMM and train it with Fastai.
▶ Export the trained model with TorchScript for inference.
▶ Create a functional Android app and run the inference on your device.

🔥 Inference time is about 100-200ms on my Pixel 3 XL!

Deploying GPT-like language models on a chatbot is tricky.

You might wonder
• How to access the model?
• Where to host the bot?

In this 🧵I walk you through how easily I deployed a GPT-J-6B model by #EleutherAI on a #Telegram bot with @huggingface and @Gradio.

For FREE 🚀 By the end of this🧵, you’ll have your very own Telegram bot that can query the GPT-J model with any text you send it 👇

Picking an object detection model is a PAIN.

Within the YOLO family there's YOLOv1 to YOLOv5, YOLOR, YOLOX, PPYOLO.. it's never ending 😵

How do you pick the right one for your application?

In this thread, I will show you how to squeze the best of YOLOX using @weights_biases By the end of this thread you will learn how to:

💡Install the Wandb client and log the YOLOX training metrics.

🔭Compare training metrics on the Wandb dashboard.

🎯 Pick the best model with mAP and FPS values.

P/S: The best model scored > 100 FPS on a CPU 🤯

Deploying object detection models on a CPU is a PAIN.

In this thread I will show you how I optimized and 10x my YOLOX model from 5 FPS to 50 FPS on a CPU.

Yes!! CPU!

And yes for FREE.

The optimized model runs FASTER on a CPU than GPU 🤯

dicksonneoh.com/portfolio/how_…

A thread 👇 By the end of this thread, you will find out how we go from this 👇🐌