티스토리 뷰
Do Batch Sizes Actually Need To Be Powers of 2?
Is the fixation on powers of 2 for efficient GPU utilization an urban myth? In this article, we explore whether this argument is true when using today's GPUs.
wandb.ai
=> 여러 batchsize 를 가지고 total runtime 을 비교함. 결과적으로, 사이즈가 2의 n 제곱이든 2의 n 제곱 + 1 이든 두드러지는 차이가 있는 것도 아니고 general 하게 적용되는 최고의 batch size 도 없음.
결론: 자기 실험이랑 gpu 에 맞는 batchsize 를 정하면 된다.
공지사항
최근에 올라온 글
최근에 달린 댓글
- Total
- Today
- Yesterday
링크
TAG
- nlp
- language model
- neurone
- GPTZero
- neural network
- q
- S
- 가우시안
- 벡터
- Pre-trained LM
- 뉴럴넷
- 워터마킹
- LM
- 뉴런
- word embedding
- Statistical Language Model
- Elmo
- cs224n
- Attention Mechanism
- Contextual Embedding
- Neural Language Model
- Bert
- 언어모델
- transformer
- weight vector
- EM
| 일 | 월 | 화 | 수 | 목 | 금 | 토 |
|---|---|---|---|---|---|---|
| 1 | ||||||
| 2 | 3 | 4 | 5 | 6 | 7 | 8 |
| 9 | 10 | 11 | 12 | 13 | 14 | 15 |
| 16 | 17 | 18 | 19 | 20 | 21 | 22 |
| 23 | 24 | 25 | 26 | 27 | 28 | 29 |
| 30 | 31 |
글 보관함