| Paper | Code | CIDEr | METEOR | BLEU-4 | SPICE | ModelName | ReleaseDate |
|---|---|---|---|---|---|---|---|
| CommonGen: A Constrained Text Generation Challenge for Generative Commonsense Reasoning | ✓ Link | 14.92 | UniLM | 2019-11-09 | |||
| TextBox 2.0: A Text Generation Library with Pre-trained Language Models | ✓ Link | 12.98 | 28.18 | 33 | BART (TextBox 2.0) | 2022-12-26 | |
| The GEM Benchmark: Natural Language Generation, its Evaluation and Metrics | 0.301 | BART | 2021-02-02 | ||||
| The GEM Benchmark: Natural Language Generation, its Evaluation and Metrics | 0.291 | T5 | 2021-02-02 |