Sentiment Evaluation Metric (F1 vs. Accuracy) #464
Description
The documentation and Hugging Face model cards for the Erlangshen sentiment analysis models reports the following results on the ASAP and ChnSentiCorp benchmarks.
| ASAP-SENT | ASAP-ASPECT | ChnSentiCorp | |
|---|---|---|---|
| Erlangshen-Roberta-110M-Sentiment | 97.77 | 97.31 | 96.61 |
| Erlangshen-Roberta-330M-Sentiment | 97.90 | 97.51 | 96.66 |
| Erlangshen-MegatronBert-1.3B-Sentiment | 98.10 | 97.80 | 97.00 |
What metric is being reported? Is it macro F1, accuracy, or some other metric?