Athletics at the 1991 Pan American Games – Women's 3000 metres
The women's 3000 metres event at the 1991 Pan American Games was held in Havana, Cuba on 4 August.[1][2][3]
Results
Rank | Name | Nationality | Time | Notes |
---|---|---|---|---|
![]() | Sabrina Dornhoefer | ![]() | 9:16.15 | |
![]() | María del Carmen Díaz | ![]() | 9:19.05 | |
![]() | Carmen Furtado | ![]() | 9:19.18 | |
4 | Lisa Harvey | ![]() | 9:19.58 | |
5 | María Luisa Servín | ![]() | 9:20.26 | |
6 | Sarah Howell | ![]() | 9:22.50 | |
7 | Milagro Rodríguez | ![]() | 9:26.55 | |
8 | Sammie Gdowski | ![]() | 9:31.64 | |
9 | Carmen Arrúa | ![]() | 9:35.03 | |
10 | Sandra Cortez | ![]() | 9:50.95 | |
11 | Yesenia Centeno | ![]() | 9:56.76 | |
12 | Bigna Samuel | ![]() | 9:57.53 | |
13 | Anna Eatherley | ![]() | 10:12.73 |
gollark: They're not deliberately making a weird pricing structure. The tokens are just a way to compact the input before it goes into the model. These things are often (partly) based on "transformers", which operate on a sequence of discrete tokens as input/output, and for which time/space complexity scales quadratically with input length. So they can't just give the thing bytes directly or something like that. And for various reasons it wouldn't make sense to give it entire words as inputs. The compromise is to break text into short tokens, which *on average* map to a certain number of words.
gollark: (not in the SCP universe, but in general, I mean)
gollark: I think that's been done a lot already. I liked https://qntm.org/ra, which is basically that.
gollark: I suppose you could argue that it isn't really relevant, since it can't run in the actual universe.
gollark: It also can't model itself.
References
This article is issued from Wikipedia. The text is licensed under Creative Commons - Attribution - Sharealike. Additional terms may apply for the media files.