I've just updated toy-library for GAN: https://t.co/t3TbkCkWbT New layers types are handled now. Don't forget to check smiley face generator also :-)
Yeah, It's built on top of @GorgoniaML 's machine learning library
#Gorgonia#MachineLearning#GAN#NeuralNetworks#DeepLearning
@kerem_gocen@GorgoniaML 1. History. I wrote Gorgonia long before things like TF or PyTorch was available.
2. No docker, no funny overheads. Compile a single binary that runs anywhere. Builds are mostly reproducible.
3. Stricter semantics (because I wrote them that way) means easier debugging
@kerem_gocen@GorgoniaML Being able to run the code anywhere. Some sort of write once, run many.
On top of that, understanding how things work and being able to add features easily (Go is easier than c++).
Last but not least: the community is small but veeeery friendly and helpful. Gorgonia works.
@VDNSeclin Hum, s/dans la région/dans l'@univ_lille... Saviez vous qu'il existait des entreprises dans cette même région dont c'est aussi le métier ? Il y a même des développeurs passionnés qui produisent du code dédié ! Exemple @GorgoniaML
@pommeDouze @stack_labs J’ai une présentation qui parle de la concurrence en Go appliquée aux calculs de tenseurs dans @GorgoniaML. Ça vous intéresse?
https://t.co/DKLtdvVDsM
@GorgoniaML’s experimental GoMachine is using context internally.
See https://t.co/EW5vK55B27
DM are open if you want to talk about it (got a present almost ready to present it)
Looking forward to contribute to @GorgoniaML v0.10.0. Meanwhile, you can test the GoMachine and/or fix the documentation :)
https://t.co/DKLtdvVDsM
#golang
Running #tinyyolov2 with @GorgoniaML GoMachine on a 4 core CPU on Darwin. The GoRoutine scheduler is using 10% of the time; I guess we can improve this. But there won't be any miracle, job needs to be done anyway.
In OSS, contributions are the oil and documentation is the soil.
@GorgoniaML has a new contributing guide for the documentation.
https://t.co/yIOuOKrGT3
First round of this talk about #DeepLearning with Gophers all the way down, and the demo has worked has expected. Thanks demo god and thanks to all people from
#golang@onnxai@KnativeProject and @GorgoniaML who makes all of this possible. ❤️