Hugging Face's logo

knowledgator
/
gliclass-x-base

Text Classification
Safetensors
GLiClass
Model card Files
xet
Community
alexandrlukashov commited on
Commit
68ef56f
·
verified ·
1 Parent(s): 1e07c00

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +24 -22
README.md CHANGED
@@ -119,26 +119,28 @@ for result in results:
119
  ### Benchmarks:
120
  Below, you can see the F1 score on several text classification datasets. All tested models were not fine-tuned on those datasets and were tested in a zero-shot setting.
121
  #### Multilingual benchmarks
122
- | Dataset | knowledgator/gliclass-x-base | knowledgator/gliclass-base-v3.0 | knowledgator/gliclass-large-v3.0 |
123
- |--------------------------|------------------------------|---------------------------------|----------------------------------|
124
- | FredZhang7/toxi-text-3M | 0.5972 | 0.5072 | 0.6118 |
125
- | SetFit/xglue_nc | 0.5014 | 0.5348 | 0.5378 |
126
- | Davlan/sib200_14classes | 0.4663 | 0.2867 | 0.3173 |
127
- | uhhlt/GermEval2017 | 0.3999 | 0.4010 | 0.4299 |
128
- | dolfsai/toxic_es | 0.1250 | 0.1399 | 0.1412 |
 
129
  #### General benchmarks
130
- | Dataset | gliclass-x-base | gliclass-base-v3.0 | gliclass-large-v3.0 |
131
- |--------------------------------|-----------------|--------------------|---------------------|
132
- | SetFit/CR | 0.8630 | 0.9398 | 0.9400 |
133
- | SetFit/sst2 | 0.8554 | 0.9192 | 0.9192 |
134
- | SetFit/sst5 | 0.3287 | 0.4606 | 0.4606 |
135
- | AmazonScience/massive | 0.2611 | 0.5649 | 0.5650 |
136
- | stanfordnlp/imdb | 0.8840 | 0.9366 | 0.9366 |
137
- | SetFit/20_newsgroups | 0.4116 | 0.5958 | 0.5958 |
138
- | SetFit/enron_spam | 0.5929 | 0.7584 | 0.7584 |
139
- | PolyAI/banking77 | 0.3098 | 0.5574 | 0.5574 |
140
- | takala/financial_phrasebank | 0.7851 | 0.9000 | 0.9000 |
141
- | ag_news | 0.6815 | 0.7181 | 0.7181 |
142
- | dair-ai/emotion | 0.3667 | 0.4506 | 0.4510 |
143
- | MoritzLaurer/cap_sotu | 0.3935 | 0.4589 | 0.6118 |
144
- | cornell-movie-review-data/rotten_tomatoes | 0.8411 | 0.8411 | 0.8411 |
 
 
119
  ### Benchmarks:
120
  Below, you can see the F1 score on several text classification datasets. All tested models were not fine-tuned on those datasets and were tested in a zero-shot setting.
121
  #### Multilingual benchmarks
122
+ | Dataset | gliclass-x-base | gliclass-base-v3.0 | gliclass-large-v3.0 |
123
+ | ------------------------ | --------------- | ------------------ | ------------------- |
124
+ | FredZhang7/toxi-text-3M | 0.5972 | 0.5072 | 0.6118 |
125
+ | SetFit/xglue\_nc | 0.5014 | 0.5348 | 0.5378 |
126
+ | Davlan/sib200\_14classes | 0.4663 | 0.2867 | 0.3173 |
127
+ | uhhlt/GermEval2017 | 0.3999 | 0.4010 | 0.4299 |
128
+ | dolfsai/toxic\_es | 0.1250 | 0.1399 | 0.1412 |
129
+ | **Average** | **0.41796** | **0.37392** | **0.4076** |
130
  #### General benchmarks
131
+ | Dataset | gliclass-x-base | gliclass-base-v3.0 | gliclass-large-v3.0 |
132
+ | ---------------------------- | --------------- | ------------------ | ------------------- |
133
+ | SetFit/CR | 0.8630 | 0.9398 | 0.9400 |
134
+ | SetFit/sst2 | 0.8554 | 0.9192 | 0.9192 |
135
+ | SetFit/sst5 | 0.3287 | 0.4606 | 0.4606 |
136
+ | AmazonScience/massive | 0.2611 | 0.5649 | 0.5650 |
137
+ | stanfordnlp/imdb | 0.8840 | 0.9366 | 0.9366 |
138
+ | SetFit/20\_newsgroups | 0.4116 | 0.5958 | 0.5958 |
139
+ | SetFit/enron\_spam | 0.5929 | 0.7584 | 0.7584 |
140
+ | PolyAI/banking77 | 0.3098 | 0.5574 | 0.5574 |
141
+ | takala/financial\_phrasebank | 0.7851 | 0.9000 | 0.9000 |
142
+ | ag\_news | 0.6815 | 0.7181 | 0.7181 |
143
+ | dair-ai/emotion | 0.3667 | 0.4506 | 0.4510 |
144
+ | MoritzLaurer/cap\_sotu | 0.3935 | 0.4589 | 0.6118 |
145
+ | cornell/rotten\_tomatoes | 0.8411 | 0.8411 | 0.8411 |
146
+ | **Average** | **0.5902** | **0.7001** | **0.7120** |