2 months ago

TODAY (March 31) at 9:00 PM CET my generative token "nd-genuary32nd-plants-chatgpt3-001" will be released on #fxhash.

The foundation was built by using #ChatGPT and this is the result of me transforming that code into a NFT.

#generativeart #genart #nd #chatgpt3 #plants #flower #sun #midnight #nerddisco #agpl #opensource #aiinfluenced #canvas2d #params #javascript #generative #nft #crypto #tezos #blockchain #genart #art #mastoart

A field of 8 colorful flowers. Drawn on a 2d canvas with JavaScript, the first one is small and purple. The second one is big and yellow/red. The third one is even bigger and green / yellow. The 4th is smaller as the other two before and purple/orange. The 5th is a bit smalleer than the one before, purple/orange. The 6th is is a special plant, has transparent petals and is purple/yellow. The 7th is even smaller as the ones before and red/purple. The last and 8th one is as small as the first one in yellow/purple. The sun is a greyish purple, the sky is greenish grey.
2 months ago

The NFT itself will be released on March 31 at 9:00 PM CET on
#fxhash and it’s limited to 32 editions because of #genuary32nd. It's also using #params, so make sure to check it out 🙏

25% will go to #TezQuakeAid

2 months ago

I'm also curious how people are using dry-validation/dry-schema/dry-struct to handle sinatra params from forms or sidekiq job arguments. Do people use the dry-* libraries directly, or do they use one of the other plugin libraries such as sinatra-validation, sinatra-dry_params, or sidekiq-dry?
#dryrb #sinatra #sidekiq #params

@simon right, my suspicion is that 3→4 may not be as much of a big step in terms of #params as 2→3 was, but there's nothing to base that off of other than the noticeable omission of it and the similar performance on some tasks.

Seems possible that the lion's share of the "more bigger-er" effect at work here is in the corpus/training rather than the architecture

📝 ZiCo: Zero-Shot NAS via Inverse Coefficient of Variation on Gradients 🧠

"ZiCo is the first training-free proxy that works consistently better than the number of network parameters (#Params), the previous SOTA training-free proxy." [gal30b+] 🤖 #LG

🔗 #arxiv

Figure 2: Training loss and Test loss vs. standard deviation of gradients for two-layer MLPs with ReLU on MNIST after one training epoch. Networks with smaller σ tend to have lower training loss and test loss values. We provide more results in Sec C.1.
Figure 1: Training loss vs. square sum of mean gradients and the sum of gradients variances for linear networks on MNIST after one epoch. Clearly, larger mean gradient values lead to lower loss values; also, networks with smaller ∑ j σ 2 j have lower loss values.
Figure 8: Real test accuracy vs. various proxies on NASBench101 search space for CIFAR10 dataset. τ and ρ are short for Kendall’s τ and Spearman’s ρ, respectively.
Figure 3: The correlation coefficients between various zero-cost proxies vs. test accuracy on NASBench101 search space for CIFAR10 dataset. As shown, our proposed ZiCo correlates best with the real test accuracy and is significantly better than all other proxies except for Zen-score.
3 years ago

クソ眠くて考え忘れてたんだけど、#params にclass外部から値ねじこんだりとかしてないよね?

3 years ago

Today I learned about URLSearchParams: A nice way to handle url parameters with vanilla javascript!

#javascript #url #params #mdn #mozilla #nativeApisBestApis