HackerTrans
トップ新着トレンドコメント過去質問紹介求人

eric15342335

no profile record

コメント

eric15342335
·4 か月前·議論
just kidding: will someone make a ublock filter list for this website?
eric15342335
·5 か月前·議論
My first impression is that the model sounds slightly more human and a little more praising. Still comparing the ability.
eric15342335
·5 か月前·議論
Interesting. Have already sent 6 emails :)
eric15342335
·5 か月前·議論
Correct me if I am wrong, I think this is related to the term "covariate shift" (change in model input distribution x) and "concept drift".
eric15342335
·8 か月前·議論
Update: it is available at https://aistudio.google.com now!
eric15342335
·9 か月前·議論
correlation is not causation
eric15342335
·9 か月前·議論
What about Pylint? iirc pylint has code duplication check as well. is it the same thing?
eric15342335
·10 か月前·議論
Not sure if I get it correctly:

They trained a thing to learn mimicking the full attention distribution but only filtering the top-k (k=2048) most important attention tokens so that when the context window increases, the compute does not go up linearly but constantly for the attention->[query,key] process (it does grow up linearly in the graph because you still need to roughly scan the entire context window (which an "indexer" will do), but just very roughly here in order to speed up things, which is O(L) here).
eric15342335
·2 年前·議論
I am a university year 2 student learning about basic mathematics and statistics related to neural networks. One thing that shocks me is that there isn't an "incremental" solution for building larger (more parameters) AI models (like GPT-4) despite having one in a smaller size e.g. GPT-3.5 (I saw the term "incremental (compiling)" nearly everywhere in the software engineering industry). I am curious how is this not possible theortically?