Skip to content

Commit ad37f56

Browse files
authored
Update README.md
1 parent 1388e98 commit ad37f56

File tree

1 file changed

+20
-0
lines changed

1 file changed

+20
-0
lines changed

README.md

Lines changed: 20 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -42,9 +42,29 @@ CUDA_VISIBLE_DEVICES=0 python run_bild_translation.py --model bild --small [smal
4242
* `[RB]` is the rollback threshold (normally 2~5 works fine). `[FB]` is the fallback threshold that can have a value from 0 to 1. For more details of these two hyperparameters, please refer to our paper.
4343

4444

45+
4546
We also provide a command for running the baseline model:
4647
```
4748
CUDA_VISIBLE_DEVICES=0 python run_bild_translation.py --model [model_path] \
4849
--dataset_name iwslt2017 --dataset_config iwslt2017-de-en --source_lang de --target_lang en
4950
```
5051
* `[model_path]` is the path to the baseline model (e.g. `[small_model_path]` or `[large_model_path]`)
52+
53+
## Pretrained Checkpoints
54+
55+
We provide finetuned checkpoints that were used for the evaluations in our paper.
56+
57+
| Dataset | Model | Link |
58+
| -------- | -------- | -------- |
59+
| IWSLT-2017-De-En | mT5-small | [link](https://huggingface.co/kssteven/mT5-small-iwslt2017-de-en) |
60+
| IWSLT-2017-De-En | mT5-small (aligned) | [link](https://huggingface.co/kssteven/mT5-small-iwslt2017-de-en-bild-aligned) |
61+
| IWSLT-2017-De-En | mT5-large | [link](https://huggingface.co/kssteven/mT5-large-iwslt2017-de-en) |
62+
| WMT-2014-De-En | mT5-small | [link](https://huggingface.co/kssteven/mT5-small-wmt2014-de-en) |
63+
| WMT-2014-De-En | mT5-small (aligned) | [link](https://huggingface.co/kssteven/mT5-small-wmt2014-de-en-bild-aligned) |
64+
| WMT-2014-De-En | mT5-large | [link](https://huggingface.co/kssteven/mT5-large-wmt2014-de-en) |
65+
| XSUM | T5-small | [link](https://huggingface.co/kssteven/T5-small-xsum) |
66+
| XSUM | T5-small (aligned) | [link](https://huggingface.co/kssteven/T5-small-xsum-bild-aligned) |
67+
| XSUM | T5-large | [link](https://huggingface.co/kssteven/T5-large-xsum) |
68+
| CNNDM | T5-small | [link](https://huggingface.co/kssteven/T5-small-cnndm) |
69+
| CNNDM | T5-small (aligned) | [link](https://huggingface.co/kssteven/T5-small-cnndm-bild-aligned) |
70+
| CNNDM | T5-large | [link](https://huggingface.co/kssteven/T5-large-cnndm) |

0 commit comments

Comments
 (0)