paper
-with-
me
Papers
Browse State-of-the-Art
Datasets
Methods
AI Agents
Trends
Digest
🌙
Cross-Lingual Question Answering
벤치마크
Cross-Lingual Question Answering on
TyDiQA-GoldP
11개 결과 ·
⬇ CSV
·
JSON
EM
42.8
52.58
62.35
72.12
81.9
2020-10
2026-09
Decoupled — 42.8 (2020-10-24)
ByT5 (fine-tuned) — 81.9 (2021-05-28)
ByT5 XXL — 60.0 (2021-05-28)
PaLM-540B (CoT) — 52.9 (2022-04-05)
U-PaLM 62B (fine-tuned) — 78.4 (2022-10-20)
Flan-U-PaLM 540B (direct-prompting) — 68.3 (2022-10-20)
Flan-PaLM 540B (direct-prompting) — 67.8 (2022-10-20)
U-PaLM-540B (CoT) — 54.6 (2022-10-20)
Decoupled — 42.8 (2020-10-24)
ByT5 (fine-tuned) — 81.9 (2021-05-28)
2020-10-24 — Decoupled: EM 42.8
2021-05-28 — ByT5 (fine-tuned): EM 81.9
Rank
Model
EM
F1
Paper
Code
Year
1
ByT5 (fine-tuned)
81.9
–
ByT5: Towards a token-free future with pre-trained byte-to-byte models
huggingface/transformers
·
google-research/byt5
·
ufal/multilexnorm2021
·
+2
2021
2
U-PaLM 62B (fine-tuned)
78.4
88.5
Transcending Scaling Laws with 0.1% Extra Compute
2022
3
Flan-U-PaLM 540B (direct-prompting)
68.3
–
Scaling Instruction-Finetuned Language Models
google-research/flan
·
declare-lab/flan-alpaca
·
formulamonks/llm-benchmarker-suite
·
+6
2022
4
Flan-PaLM 540B (direct-prompting)
67.8
–
Scaling Instruction-Finetuned Language Models
google-research/flan
·
declare-lab/flan-alpaca
·
formulamonks/llm-benchmarker-suite
·
+6
2022
5
ByT5 XXL
60.0
75.3
ByT5: Towards a token-free future with pre-trained byte-to-byte models
huggingface/transformers
·
google-research/byt5
·
ufal/multilexnorm2021
·
+2
2021
6
U-PaLM-540B (CoT)
54.6
–
Transcending Scaling Laws with 0.1% Extra Compute
2022
7
PaLM-540B (CoT)
52.9
–
PaLM: Scaling Language Modeling with Pathways
lucidrains/CoCa-pytorch
·
lucidrains/PaLM-pytorch
·
google/paxml
·
+4
2022
8
Decoupled
42.8
58.1
Rethinking embedding coupling in pre-trained language models
PaddlePaddle/PaddleNLP
·
yangyucheng000/University
·
yangyucheng000/University
·
+1
2020
9
PaLM 2-L (one-shot)
–
73.6
PaLM 2 Technical Report
eternityyw/tram-benchmark
2023
10
PaLM 2-S (one-shot)
–
73.3
PaLM 2 Technical Report
eternityyw/tram-benchmark
2023
11
PaLM 2-M (one-shot)
–
73.3
PaLM 2 Technical Report
eternityyw/tram-benchmark
2023
1–11 / 11
페이지당
10
20
50
100