jafermarq

6 ms·

1.
by jafermarq 2y ago
Do you mean this example? https://github.com/adap/flower/tree/main/examples/quickstart...
2.
by jafermarq 2y ago
Amazing to see the new docs page about Flower Network Communication: https://flower.ai/docs/framework/ref-flower-network-communic...
3.
Show HN: Federation of robots collaboratively train an object manipulation model(github.com)
7 points by jafermarq 2y ago|0 comments
4.
by jafermarq 3y ago
In a world starving for GPUs, we need to start looking for alternatives. Luckily embedded processors keep getting both faster and cheaper -- making them a increasingly attractive for ML. In this blog, we benchmark the Raspberry Pi 5 for on-
5.
Raspberry Pi 5: Ready for Federated Vision(flower.dev)
8 points by jafermarq 3y ago|1 comments
6.
by jafermarq 3y ago
If the task requires labels yes. Alternatively, an auxiliary model (not part of the training) can generate pseudo labels and use those during training. A more general approach (for which a few works in FL have proven to work pretty well) wo
7.
by jafermarq 3y ago
Yes. The example uses Whisper-tiny which is 39M, a perfect match for the downstream task of keyword spotting. Just one line needs to be changed in the code to run a larger Whisper model :)
8.
by jafermarq 3y ago
yes. Finetuning a whisper model on a RPi 5 is ~2x faster than on the RPi 4. Other stages involving data pre-processing with HF dataset is again 2x-3x faster.
9.
by jafermarq 3y ago
yeah. with FL it should be possible to make sense out of all data that is distributed across devices without ever having to move it to a central location (i.e. collect it). In the case of speech data, users participating in a federated sett