vocabtrimmer/xlm-roberta-base-trimmed-de-5000-tweet-sentiment-de
This model is a fine-tuned version of /home/asahiushio/Projects/lm-vocab-trimmer/ckpts/xlm-roberta-base-trimmed-de-5000 on the
cardiffnlp/tweet_sentiment_multilingual (german).
Following metrics are computed on the test split of
cardiffnlp/tweet_sentiment_multilingual(german).
| eval_f1_micro | eval_recall_micro | eval_precision_micro | eval_f1_macro | eval_recall_macro | eval_precision_macro | eval_accuracy | |
|---|---|---|---|---|---|---|---|
| 0 | 70.57 | 70.57 | 70.57 | 70.63 | 70.57 | 71.01 | 70.57 |
Check the result file here.
