HackerTrans
TopNewTrendsCommentsPastAskShowJobs

danieljanes

no profile record

Submissions

Flower Labs and Starcloud Reach a Major AI Milestone in Orbit

flower.ai
6 points·by danieljanes·5개월 전·1 comments

Flower 1.12 release (open-source decentralized AI)

flower.ai
7 points·by danieljanes·2년 전·0 comments

Flower 1.11: Auto-deploy code, isolate ClientApp execution

flower.ai
9 points·by danieljanes·2년 전·0 comments

New short course (2/2): Federated Fine-tuning of LLMs on Private Data

learn.deeplearning.ai
11 points·by danieljanes·2년 전·0 comments

New short course (1/2): Intro to Federated (Machine) Learning

deeplearning.ai
10 points·by danieljanes·2년 전·0 comments

Flower 1.10

discuss.flower.ai
7 points·by danieljanes·2년 전·0 comments

Flower 1.9 (a friendly federated learning framework)

discuss.flower.ai
8 points·by danieljanes·2년 전·0 comments

Flower 1.8

flower.ai
8 points·by danieljanes·2년 전·0 comments

Flower 1.7 (train AI on distributed data)

flower.dev
5 points·by danieljanes·2년 전·0 comments

Federated finetuning of Whisper on Raspberry Pi 5

flower.dev
90 points·by danieljanes·3년 전·20 comments

comments

danieljanes
·3개월 전·discuss
We're not "still" using Markdown, we're only getting started.

Markdown is only going to get more popular as AI agents usage grows.
danieljanes
·5개월 전·discuss
Flower Labs and Starcloud are sharing a major milestone: the successful execution of a decentralized AI workload using Flower on an operational Starcloud satellite.
danieljanes
·5개월 전·discuss
ViT (Vision Transformer) fine-tuned on a Starcloud satellite in space (using the Flower framework) -- to the best of our knowledge, this is a world first: https://flower.ai/blog/2026-02-02-flower-labs-and-starcloud-...
danieljanes
·2년 전·discuss
Not having to install CUDA is a killer feature, looking forward to DGX OS
danieljanes
·2년 전·discuss
Given the code quality and rigid testing, SQLite is probably the last project that should be rewritten. It'd be great to see all other C code rewritten first!
danieljanes
·2년 전·discuss
We use Pyenv successfully for developing the Flower open-source project. We use a few simple Bash scripts to manage virtual environments with different Python versions via pyenv and the pyenv-virtualenv plugin.

The main scripts are `venv-create.sh`, `venv-delete.sh` and `bootstrap.sh`. `venv-reset.sh` pulls these three scripts together to make reinstalling your venv a single command.

Here's the link if anyone is interested: https://github.com/adap/flower/tree/main/dev
danieljanes
·3년 전·discuss
The big opportunity on the edge is access to more data. Especially with the rise of end-to-end encryption, applications will be able to use more (and more diverse) data on the edge to get better model performance. It's generally true that training on beefier infrastructure is easier, but in the long run, nothing can beat access to better data. And edge hardware has gotten a lot faster over the last few years.
danieljanes
·3년 전·discuss
I can confirm that we're seeing 2x to 3x faster (RPi 4 vs RPi 5) in some of our early tests
danieljanes
·3년 전·discuss
One of the Flower maintainers here. The code example is primarily meant as a demonstrator to show that it's possible to fine-tune these models in a federated way on devices as small as a Raspberry Pi 5.

The bigger takeaway is that we're close to being able to train/fine-tune models with much better performance by accessing vastly more data on the edge, in a federated way.
danieljanes
·3년 전·discuss
One of the Flower maintainers here, we're planning to follow up with a more in-depth performance comparison soon
danieljanes
·3년 전·discuss
Does GGML support training on the edge? We're especially interested in training support for Android+iOS