bloomz-7b1-p3/evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/merged.csv

1	dataset	prompt	metric	value
2	xnli_ar	GPT-3 style_armt	accuracy	0.3333333333333333
3	xnli_ar	MNLI crowdsource_armt	accuracy	0.42891566265060244
4	xnli_ar	can we infer_armt	accuracy	0.3353413654618474
5	xnli_ar	guaranteed/possible/impossible_armt	accuracy	0.3755020080321285
6	xnli_ar	justified in saying_armt	accuracy	0.3349397590361446
7	xnli_ar	median	accuracy	0.3353413654618474
8	xnli_es	GPT-3 style_esmt	accuracy	0.5220883534136547
9	xnli_es	MNLI crowdsource_esmt	accuracy	0.4847389558232932
10	xnli_es	can we infer_esmt	accuracy	0.3333333333333333
11	xnli_es	guaranteed/possible/impossible_esmt	accuracy	0.3449799196787149
12	xnli_es	justified in saying_esmt	accuracy	0.3333333333333333
13	xnli_es	median	accuracy	0.3449799196787149
14	xnli_fr	GPT-3 style_frmt	accuracy	0.4791164658634538
15	xnli_fr	MNLI crowdsource_frmt	accuracy	0.3333333333333333
16	xnli_fr	can we infer_frmt	accuracy	0.42248995983935744
17	xnli_fr	guaranteed/possible/impossible_frmt	accuracy	0.41847389558232934
18	xnli_fr	justified in saying_frmt	accuracy	0.378714859437751
19	xnli_fr	median	accuracy	0.41847389558232934
20	xnli_hi	GPT-3 style_himt	accuracy	0.3389558232931727
21	xnli_hi	MNLI crowdsource_himt	accuracy	0.3333333333333333
22	xnli_hi	can we infer_himt	accuracy	0.3542168674698795
23	xnli_hi	guaranteed/possible/impossible_himt	accuracy	0.3353413654618474
24	xnli_hi	justified in saying_himt	accuracy	0.39879518072289155
25	xnli_hi	median	accuracy	0.3389558232931727
26	xnli_sw	GPT-3 style_swmt	accuracy	0.3333333333333333
27	xnli_sw	MNLI crowdsource_swmt	accuracy	0.3333333333333333
28	xnli_sw	can we infer_swmt	accuracy	0.334136546184739
29	xnli_sw	guaranteed/possible/impossible_swmt	accuracy	0.3236947791164659
30	xnli_sw	justified in saying_swmt	accuracy	0.3321285140562249
31	xnli_sw	median	accuracy	0.3333333333333333
32	xnli_ur	GPT-3 style_urmt	accuracy	0.3751004016064257
33	xnli_ur	MNLI crowdsource_urmt	accuracy	0.3751004016064257
34	xnli_ur	can we infer_urmt	accuracy	0.329718875502008
35	xnli_ur	guaranteed/possible/impossible_urmt	accuracy	0.3337349397590361
36	xnli_ur	justified in saying_urmt	accuracy	0.3285140562248996
37	xnli_ur	median	accuracy	0.3337349397590361
38	xnli_vi	GPT-3 style_vimt	accuracy	0.3333333333333333
39	xnli_vi	MNLI crowdsource_vimt	accuracy	0.3333333333333333
40	xnli_vi	can we infer_vimt	accuracy	0.342570281124498
41	xnli_vi	guaranteed/possible/impossible_vimt	accuracy	0.3333333333333333
42	xnli_vi	justified in saying_vimt	accuracy	0.3365461847389558
43	xnli_vi	median	accuracy	0.3333333333333333
44	xnli_zh	GPT-3 style_zhmt	accuracy	0.3606425702811245
45	xnli_zh	MNLI crowdsource_zhmt	accuracy	0.39598393574297186
46	xnli_zh	can we infer_zhmt	accuracy	0.351004016064257
47	xnli_zh	guaranteed/possible/impossible_zhmt	accuracy	0.3473895582329317
48	xnli_zh	justified in saying_zhmt	accuracy	0.3409638554216867
49	xnli_zh	median	accuracy	0.351004016064257
50	multiple	average	multiple	0.348644578313253