-
Notifications
You must be signed in to change notification settings - Fork 0
Files
/
Copy pathmodel_info_en.csv
Latest commit
29 lines (29 loc) · 1.58 KB
/
model_info_en.csv
1 | model_name | n_parameters | n_layers | n_hid | ppl | hellaswag | |
---|---|---|---|---|---|---|---|
2 | 0 | gpt2 | 124439808 | 12 | 768 | 25.1879940032959 | 31.53 |
3 | 1 | gpt2-medium | 354823168 | 24 | 1024 | 18.4738788604736 | 39.38 |
4 | 2 | gpt2-large | 774030080 | 36 | 1280 | 16.4541091918945 | 45.62 |
5 | 3 | gpt2-xl | 1557611200 | 48 | 1600 | 14.7950658798218 | 50.89 |
6 | 4 | opt-125m | 125239296 | 12 | 768 | 23.9556674957275 | 31.47 |
7 | 5 | opt-350m | 331196416 | 24 | 1024 | 18.8445243835449 | 36.73 |
8 | 6 | opt-1.3b | 1315758080 | 24 | 2048 | 12.6810073852539 | 54.53 |
9 | 7 | opt-2.7b | 2651596800 | 32 | 2560 | 10.8203382492065 | 61.43 |
10 | 8 | opt-6.7b | 6658473984 | 32 | 4096 | 9.43136215209961 | 68.66 |
11 | 9 | opt-13b | 12853473280 | 40 | 5120 | 8.81267166137695 | 71.2 |
12 | 10 | Llama-2-7b-hf | 6738415616 | 32 | 4096 | 4.90231943130493 | 78.59 |
13 | 11 | Llama-2-13b-hf | 13015864320 | 40 | 5120 | 4.38781452178955 | 82.13 |
14 | 12 | Qwen1.5-0.5B | 463987712 | 24 | 1024 | 13.2299013137817 | 49.05 |
15 | 13 | Qwen1.5-1.8B | 1836828672 | 24 | 2048 | 10.2931909561157 | 61.42 |
16 | 14 | Qwen1.5-4B | 3950369280 | 40 | 2560 | 8.04801464080811 | 71.58 |
17 | 15 | Qwen1.5-7B | 7721324544 | 32 | 4096 | 7.09468841552734 | 78.51 |
18 | 16 | Qwen1.5-14B | 14167290880 | 40 | 5120 | 6.6493878364563 | 81.08 |
19 | 17 | gemma-2b | 2506172416 | 18 | 2048 | 9.02534294128418 | 71.65 |
20 | 18 | gemma-7b | 8537680896 | 28 | 3072 | 6.15309047698975 | 82.47 |
21 | 19 | stablelm-2-1_6b | 1644515328 | 24 | 2048 | 7.94636297225952 | 70.45 |
22 | 20 | stablelm-3b-4e1t | 2795443200 | 32 | 2560 | 7.40775728225708 | 75.94 |
23 | 21 | stablelm-2-12b | 12143185920 | 40 | 5120 | 5.33592987060547 | 84.33 |
24 | 22 | Mistral-7B-v0.1 | 7241732096 | 32 | 4096 | 4.74079084396362 | 83.31 |
25 | 23 | mamba-130m-hf | 129135360 | 24 | 768 | 18.6168785095215 | 35.3 |
26 | 24 | mamba-370m-hf | 371516416 | 48 | 1024 | 12.92933177948 | 46.5 |
27 | 25 | mamba-790m-hf | 793204224 | 48 | 1536 | 10.8047342300415 | 55.1 |
28 | 26 | mamba-1.4b-hf | 1372178432 | 48 | 2048 | 9.76564407348633 | 59.1 |
29 | 27 | mamba-2.8b-hf | 2768345600 | 64 | 2560 | 8.56341934204102 | 66.1 |