Views
No views yet
epoch=5-step=100, global step 100, validation Kendall τ = 0.347).| Parameter | Value |
|---|---|
| Base model | xlm-roberta-large |
| Finetuning data | Amazon Bio-MQM (dev splits) |
| Training epochs | 6 (Lightning epoch index 0–5) |
| Checkpoint | epoch=5-step=100, global step 100 |
| Validation Kendall τ | 0.347 |
| Language pairs | de↔en, es↔en, fr↔en, ru↔en, zh↔en |
| Loss | mse |
| Encoder LR | 5e-07 |
| Head LR | 1e-05 |
| Batch size | 8 |
| Run | comet-bio-mqm-4yeqp7cn |
1from comet import download_model, load_from_checkpoint
2
3model_path = download_model("AdleBenSalem/comet-bio-mqm-4yeqp7cn")
4model = load_from_checkpoint(model_path)
5
6data = [{
7 "src": "The patient was administered 500 mg of amoxicillin.",
8 "mt": "Der Patient erhielt 500 mg Amoxicillin.",
9 "ref": "Dem Patienten wurden 500 mg Amoxicillin verabreicht.",
10}]
11output = model.predict(data, batch_size=8, gpus=1)
12print(output.scores)1@inproceedings{rei-etal-2020-comet,
2 title = {COMET: A Neural Framework for MT Evaluation},
3 author = {Rei, Ricardo and Stewart, Craig and Farinha, Ana C and Lavie, Alon},
4 booktitle = {Proceedings of EMNLP 2020},
5}
6
7@inproceedings{bio-mqm-2024,
8 title = {Fine-Tuned Machine Translation Metrics Struggle in Unseen Domains},
9 booktitle = {Proceedings of ACL 2024},
10}