BERTScore Default Layer Performance on WMT16

	A	B	C	D	E
1	Model	Best Layer	WMT16 To-English Pearson Correlation	Rank	Max Length
2	bert-base-uncased	9	0.6925	51	510
3	bert-large-uncased	18	0.7210	28	510
4	bert-base-cased-finetuned-mrpc	9	0.6722	74	510
5	bert-base-multilingual-cased	9	0.6681	81	510
6	roberta-base	10	0.7063	38	510
7	roberta-large	17	0.7386	20	510
8	roberta-large-mnli	19	0.7536	6	510
9	roberta-base-openai-detector	7	0.7048	41	510
10	roberta-large-openai-detector	15	0.7463	14	510
11	xlnet-base-cased	5	0.6630	86	1000000000000
12	xlnet-large-cased	7	0.6599	87	1000000000000
13	xlm-mlm-en-2048	6	0.6513	103	510
14	xlm-mlm-100-1280	10	0.6475	108	510
15	scibert-scivocab-uncased	8	0.6590	89	510
16	scibert-scivocab-cased	9	0.6536	98	510
17	scibert-basevocab-uncased	9	0.6749	72	510
18	scibert-basevocab-cased	9	0.6525	102	510
19	distilroberta-base	5	0.6798	63	510
20	distilbert-base-uncased	5	0.6757	68	510
21	distilbert-base-uncased-distilled-squad	4	0.6718	75	510
22	distilbert-base-multilingual-cased	5	0.6178	126	510
23	albert-base-v1	10	0.6542	97	510
24	albert-large-v1	17	0.6756	69	510
25	albert-xlarge-v1	16	0.7032	43	510
26	albert-xxlarge-v1	8	0.7509	10	510
27	albert-base-v2	9	0.6682	80	510
28	albert-large-v2	14	0.7009	44	510
29	albert-xlarge-v2	13	0.7317	23	510
30	albert-xxlarge-v2	8	0.7505	11	510
31	xlm-roberta-base	9	0.6507	104	510
32	xlm-roberta-large	17	0.6942	48	510
33	google/electra-small-generator	9	0.6659	82	510
34	google/electra-small-discriminator	11	0.6535	100	510
35	google/electra-base-generator	10	0.6730	73	510
36	google/electra-base-discriminator	9	0.7032	42	510
37	google/electra-large-generator	18	0.6813	61	510
38	google/electra-large-discriminator	14	0.6897	55	510
39	google/bert_uncased_L-2_H-128_A-2	1	0.5888	133	510
40	google/bert_uncased_L-2_H-256_A-4	1	0.6115	130	510
41	google/bert_uncased_L-2_H-512_A-8	1	0.6177	127	510
42	google/bert_uncased_L-2_H-768_A-12	2	0.6191	123	510
43	google/bert_uncased_L-4_H-128_A-2	3	0.6076	131	510
44	google/bert_uncased_L-4_H-256_A-4	3	0.6205	121	510
45	google/bert_uncased_L-4_H-512_A-8	3	0.6375	117	510
46	google/bert_uncased_L-4_H-768_A-12	3	0.6562	94	510
47	google/bert_uncased_L-6_H-128_A-2	5	0.6200	122	510
48	google/bert_uncased_L-6_H-256_A-4	5	0.6278	120	510
49	google/bert_uncased_L-6_H-512_A-8	5	0.6420	113	510
50	google/bert_uncased_L-6_H-768_A-12	5	0.6762	66	510
51	google/bert_uncased_L-8_H-128_A-2	7	0.6187	124	510
52	google/bert_uncased_L-8_H-256_A-4	7	0.6448	110	510
53	google/bert_uncased_L-8_H-512_A-8	6	0.6490	105	510
54	google/bert_uncased_L-8_H-768_A-12	7	0.6705	77	510
55	google/bert_uncased_L-10_H-128_A-2	8	0.6127	129	510
56	google/bert_uncased_L-10_H-256_A-4	8	0.6376	116	510
57	google/bert_uncased_L-10_H-512_A-8	9	0.6579	93	510
58	google/bert_uncased_L-10_H-768_A-12	8	0.6861	56	510
59	google/bert_uncased_L-12_H-128_A-2	10	0.6184	125	510
60	google/bert_uncased_L-12_H-256_A-4	11	0.6374	118	510
61	google/bert_uncased_L-12_H-512_A-8	10	0.6588	91	510
62	google/bert_uncased_L-12_H-768_A-12	9	0.6759	67	510
63	amazon/bort	0	0.4193	140	510
64	facebook/bart-base	6	0.7122	32	1022
65	facebook/bart-large	10	0.7449	15	1022
66	facebook/bart-large-cnn	10	0.7393	19	1022
67	facebook/bart-large-mnli	11	0.7532	7	1022
68	facebook/bart-large-xsum	9	0.7496	13	1022
69	t5-small	6	0.6814	60	511
70	t5-base	11	0.7096	34	511
71	t5-large	23	0.7244	27	511
72	vinai/bertweet-base	9	0.6529	101	126
73	microsoft/deberta-base	9	0.7088	36	510
74	microsoft/deberta-base-mnli	9	0.7395	17	510
75	microsoft/deberta-large	16	0.7512	9	510
76	microsoft/deberta-large-mnli	18	0.7736	3	510
77	microsoft/deberta-xlarge	18	0.7569	5	510
78	microsoft/deberta-xlarge-mnli	40	0.7781	1	510
79	YituTech/conv-bert-base	10	0.7058	39	510
80	YituTech/conv-bert-small	10	0.6544	95	510
81	YituTech/conv-bert-medium-small	9	0.6590	90	510
82	microsoft/mpnet-base	8	0.7250	26	510
83	squeezebert/squeezebert-uncased	9	0.6544	96	510
84	squeezebert/squeezebert-mnli	9	0.6655	83	510
85	squeezebert/squeezebert-mnli-headless	9	0.6655	83	510
86	tuner007/pegasus_paraphrase	15	0.7188	30	59
87	google/pegasus-large	8	0.6396	115	1023
88	google/pegasus-xsum	11	0.6837	58	511
89	sshleifer/tiny-mbart	2	0.0282	141	1022
90	facebook/mbart-large-cc25	12	0.6583	92	1022
91	facebook/mbart-large-50	12	0.6465	109	1022
92	facebook/mbart-large-en-ro	12	0.6791	64	1022
93	facebook/mbart-large-50-many-to-many-mmt	12	0.6904	53	1022
94	facebook/mbart-large-50-one-to-many-mmt	12	0.6848	57	1022
95	allenai/led-base-16384	6	0.7122	31	16382
96	facebook/blenderbot_small-90M	7	0.6489	106	512
97	facebook/blenderbot-400M-distill	2	0.5875	134	127
98	microsoft/prophetnet-large-uncased	4	0.5865	135	511
99	microsoft/prophetnet-large-uncased-cnndm	7	0.6478	107	511
100	microsoft/xprophetnet-large-wiki100-cased	7	0.5852	136	511