Evaluation results for ibm/ColD-Fusion-bert-base-uncased-itr23-seed0 model as a base model for other tasks
#1
by
eladven
- opened
README.md
CHANGED
|
@@ -51,6 +51,20 @@ output = model(encoded_input)
|
|
| 51 |
```
|
| 52 |
|
| 53 |
## Evaluation results
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 54 |
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
|
| 55 |
When fine-tuned on downstream tasks, this model achieves the following results:
|
| 56 |
|
|
|
|
| 51 |
```
|
| 52 |
|
| 53 |
## Evaluation results
|
| 54 |
+
|
| 55 |
+
## Model Recycling
|
| 56 |
+
|
| 57 |
+
[Evaluation on 36 datasets](https://ibm.github.io/model-recycling/model_gain_chart?avg=3.44&mnli_lp=nan&20_newsgroup=2.07&ag_news=-0.46&amazon_reviews_multi=0.34&anli=2.14&boolq=5.42&cb=12.41&cola=0.15&copa=8.55&dbpedia=0.04&esnli=1.02&financial_phrasebank=15.57&imdb=0.52&isear=0.22&mnli=0.65&mrpc=5.02&multirc=-0.61&poem_sentiment=18.89&qnli=-0.60&qqp=0.29&rotten_tomatoes=4.55&rte=18.00&sst2=2.18&sst_5bins=2.72&stsb=2.71&trec_coarse=1.14&trec_fine=12.67&tweet_ev_emoji=0.28&tweet_ev_emotion=1.16&tweet_ev_hate=2.20&tweet_ev_irony=0.61&tweet_ev_offensive=-0.37&tweet_ev_sentiment=0.82&wic=2.58&wnli=1.55&wsc=0.38&yahoo_answers=-1.02&model_name=ibm%2FColD-Fusion-bert-base-uncased-itr23-seed0&base_name=bert-base-uncased) using ibm/ColD-Fusion-bert-base-uncased-itr23-seed0 as a base model yields average score of 75.64 in comparison to 72.20 by bert-base-uncased.
|
| 58 |
+
|
| 59 |
+
The model is ranked 1st among all tested models for the bert-base-uncased architecture as of 09/01/2023
|
| 60 |
+
Results:
|
| 61 |
+
|
| 62 |
+
| 20_newsgroup | ag_news | amazon_reviews_multi | anli | boolq | cb | cola | copa | dbpedia | esnli | financial_phrasebank | imdb | isear | mnli | mrpc | multirc | poem_sentiment | qnli | qqp | rotten_tomatoes | rte | sst2 | sst_5bins | stsb | trec_coarse | trec_fine | tweet_ev_emoji | tweet_ev_emotion | tweet_ev_hate | tweet_ev_irony | tweet_ev_offensive | tweet_ev_sentiment | wic | wnli | wsc | yahoo_answers |
|
| 63 |
+
|---------------:|----------:|-----------------------:|--------:|--------:|--------:|--------:|-------:|----------:|--------:|-----------------------:|-------:|--------:|--------:|--------:|----------:|-----------------:|--------:|--------:|------------------:|--------:|--------:|------------:|--------:|--------------:|------------:|-----------------:|-------------------:|----------------:|-----------------:|---------------------:|---------------------:|--------:|--------:|------:|----------------:|
|
| 64 |
+
| 85.1168 | 89.1333 | 66.26 | 49.0938 | 74.3731 | 76.7857 | 81.9751 | 58 | 78.2 | 90.7268 | 84.1 | 92.096 | 69.296 | 84.3775 | 87.0098 | 59.3647 | 85.5769 | 89.2733 | 90.5639 | 89.3996 | 77.9783 | 94.1514 | 55.5204 | 88.5727 | 97.2 | 81 | 36.282 | 81.0697 | 55.0505 | 68.3673 | 85 | 70.3028 | 65.8307 | 52.1127 | 62.5 | 71.3 |
|
| 65 |
+
|
| 66 |
+
|
| 67 |
+
For more information, see: [Model Recycling](https://ibm.github.io/model-recycling/)
|
| 68 |
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
|
| 69 |
When fine-tuned on downstream tasks, this model achieves the following results:
|
| 70 |
|