HackerTrans
TopNewTrendsCommentsPastAskShowJobs

jumpCastle

no profile record

comments

jumpCastle
·작년·discuss
It was a good benchmark until it entered the training set.
jumpCastle
·작년·discuss
But open source like aider
jumpCastle
·작년·discuss
Deepseek v3 was FP8
jumpCastle
·작년·discuss
It's quite similar to muP

https://github.com/microsoft/mup
jumpCastle
·2년 전·discuss
Stock prices cannot go up without if the planet is destroyed
jumpCastle
·2년 전·discuss
Attention was invented because Bengio lab had to be disciplined about a black box (google had more compute)
jumpCastle
·2년 전·discuss
Aren't services like runpod solve half of these concerns?
jumpCastle
·2년 전·discuss
Are the updates really slow?