Google engineer explained how to fine-tune a tiny LLM from 46% to 90% accuracy on your phone in 21 minutes - better than $1500 on-device AI bootcamps.
pick Gemma 270M -> generate synthetic task data -> fine-tune with LoRA -> quantize to int4 -> deploy to Pixel and hit 2000 tokens per second.
That loop is how a 270M model beats a 70B one on your task, running fully offline in your pocket.
Gemma 270M + synthetic data + LoRA + int4 quantization + on-device runtime - that's the stack.
Watch and save it, then fine-tune your own tiny agent tonight.
@kathmandupost People are dying in hospitals while this government talks about reform. Corruption has spread through every institution under their rule, hollowing out healthcare, looting public funds, and protecting incompetence.
@NepaliTimes Why don't you ask your different fathers who were in the govt who looted many many decades one killed 17k other do human trafficking and other many corruption scandle
Do you think this is india doing godi media work.
@NepaliTimes When your income is ads from the same old corrupt network,
you’ll obviously defame the new voice.
When you can’t speak against Oli–Deuba–Prachanda because that’s where the money comes from 💰🤡
Now their media dogs are barking.
The louder they bark, the closer their end.
@getpeid Issue: The Nothing Phone (2a) lacks crucial battery notifications.
Low Battery Notification: There is no pop-up or alert when the battery falls below 20% & warning or timer, unlike Xiaomi phones which display a "Shutting down in 60 seconds" timer.
https://t.co/QbKyFtkzYH