1930–31 Scottish Division Two
The 1930–31 Scottish Second Division was won by Third Lanark who, along with second placed Dundee United, were promoted to the First Division. Bo'ness finished bottom.
Season | 1930–31 |
---|---|
Champions | Third Lanark |
Promoted | Third Lanark Dundee United |
← 1929–30 1931–32 → |
Table
Pos | Team | Pld | W | D | L | GF | GA | GD | Pts | Promotion or relegation |
---|---|---|---|---|---|---|---|---|---|---|
1 | Third Lanark | 38 | 27 | 7 | 4 | 107 | 42 | +65 | 61 | Promotion to the 1931–32 First Division |
2 | Dundee United | 38 | 21 | 8 | 9 | 93 | 54 | +39 | 50 | |
3 | Dunfermline Athletic | 38 | 20 | 7 | 11 | 83 | 50 | +33 | 47 | |
4 | Raith Rovers | 38 | 20 | 6 | 12 | 93 | 72 | +21 | 46 | |
5 | St Johnstone | 38 | 19 | 6 | 13 | 76 | 61 | +15 | 44 | |
6 | Queen of the South | 38 | 18 | 6 | 14 | 83 | 66 | +17 | 42 | |
7 | East Stirlingshire | 38 | 17 | 7 | 14 | 85 | 74 | +11 | 41 | |
8 | Montrose | 38 | 19 | 3 | 16 | 75 | 90 | −15 | 41 | |
9 | Albion Rovers | 38 | 14 | 11 | 13 | 83 | 84 | −1 | 39 | |
10 | Dumbarton | 38 | 15 | 8 | 15 | 73 | 72 | +1 | 38 | |
11 | St Bernard's | 38 | 14 | 9 | 15 | 85 | 66 | +19 | 37 | |
12 | Forfar Athletic | 38 | 15 | 6 | 17 | 80 | 84 | −4 | 36 | |
13 | Alloa Athletic | 38 | 15 | 5 | 18 | 65 | 87 | −22 | 35 | |
14 | King's Park | 38 | 14 | 6 | 18 | 78 | 70 | +8 | 34 | |
15 | Arbroath | 38 | 15 | 4 | 19 | 83 | 94 | −11 | 34 | |
16 | Brechin City | 38 | 13 | 7 | 18 | 52 | 84 | −32 | 33 | |
17 | Stenhousemuir | 38 | 12 | 6 | 20 | 75 | 101 | −26 | 30 | |
18 | Armadale | 38 | 13 | 2 | 23 | 74 | 99 | −25 | 28 | |
19 | Clydebank | 38 | 10 | 2 | 26 | 61 | 108 | −47 | 22 | |
20 | Bo'ness | 38 | 9 | 4 | 25 | 54 | 100 | −46 | 22 |
Source:
gollark: For more than a minute.
gollark: Int8 apparently causes it to just output random noise and I never got round to trying quantisation aware training for it.
gollark: It's quite strange that apparently BERT can be statically quantized without any extra training and retains decent accuracy but GPT-Neo emits nonsense going through the same process.
gollark: I was looking into quantization-aware training a while ago, but on the 125M model, and running that for a bit made it produce English-looking nonsense instead of random noise.
gollark: I think there's technically a way to swap bits of the model in and out of VRAM but it would still be quite slow.
This article is issued from Wikipedia. The text is licensed under Creative Commons - Attribution - Sharealike. Additional terms may apply for the media files.