gonglinyuan commited on
Commit
fabbe38
1 Parent(s): 1245691

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +10 -2
README.md CHANGED
@@ -119,7 +119,7 @@ Official repository: https://github.com/gonglinyuan/metro_t0
119
 
120
  # METRO-T0
121
 
122
- Paper: Model-Generated Pretraining Signals Improves Zero-Shot Generalization of Text-to-Text Transformers (TODO) (ACL 2023)
123
 
124
  METRO-T0 is a T5-style text-to-text Transformer pretrained using model-generated pretraining signals, prompt-finetuned on a family of public NLP tasks proposed in [T0](https://arxiv.org/abs/2110.08207).
125
  METRO-T0 is highly parameter efficient. For example, METRO-T0-Large++ (775M parameters) outperforms GPT-3 (175B parameters) and T0-3B (3B parameters) on a wide range of NLP tasks.
@@ -165,5 +165,13 @@ print(tokenizer.decode(outputs[0], skip_special_tokens=True)) # expected: posit
165
  If you find the code and models useful for your research, please cite the following paper:
166
 
167
  ```
168
- TODO
 
 
 
 
 
 
 
 
169
  ```
 
119
 
120
  # METRO-T0
121
 
122
+ Paper: [Model-Generated Pretraining Signals Improves Zero-Shot Generalization of Text-to-Text Transformers](https://arxiv.org/abs/2305.12567) (ACL 2023)
123
 
124
  METRO-T0 is a T5-style text-to-text Transformer pretrained using model-generated pretraining signals, prompt-finetuned on a family of public NLP tasks proposed in [T0](https://arxiv.org/abs/2110.08207).
125
  METRO-T0 is highly parameter efficient. For example, METRO-T0-Large++ (775M parameters) outperforms GPT-3 (175B parameters) and T0-3B (3B parameters) on a wide range of NLP tasks.
 
165
  If you find the code and models useful for your research, please cite the following paper:
166
 
167
  ```
168
+ @misc{gong2023modelgenerated,
169
+ title={Model-Generated Pretraining Signals Improves Zero-Shot Generalization of Text-to-Text Transformers},
170
+ author={Linyuan Gong and Chenyan Xiong and Xiaodong Liu and Payal Bajaj and Yiqing Xie and Alvin Cheung and Jianfeng Gao and Xia Song},
171
+ year={2023},
172
+ eprint={2305.12567},
173
+ archivePrefix={arXiv},
174
+ primaryClass={cs.CL},
175
+ url={https://arxiv.org/abs/2305.12567}
176
+ }
177
  ```