Hugging Face's logo

poltextlab
/
xlm-roberta-large-german-cap

Text Classification
Transformers
PyTorch
German
xlm-roberta
zero-shot-classification
Inference Endpoints
Model card Files Community
1
poltextlab commited on
Commit
263e40a
•
1 Parent(s): a8b1cd0

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +26 -26
README.md CHANGED
@@ -90,34 +90,34 @@ training_args = TrainingArguments(
90
  We also incorporated an EarlyStoppingCallback in the process with a patience of 2 epochs.
91
 
92
  ## Model performance
93
- The model was evaluated on a test set of 5227 examples (10% of the available data).<br>
94
- Model accuracy is **0.7**.
95
  | label | precision | recall | f1-score | support |
96
  |:-------------|------------:|---------:|-----------:|----------:|
97
- | 0 | 0.62 | 0.65 | 0.63 | 510 |
98
- | 1 | 0.73 | 0.67 | 0.7 | 406 |
99
- | 2 | 0.8 | 0.83 | 0.81 | 239 |
100
- | 3 | 0.75 | 0.78 | 0.76 | 136 |
101
- | 4 | 0.72 | 0.65 | 0.68 | 386 |
102
- | 5 | 0.79 | 0.82 | 0.81 | 331 |
103
- | 6 | 0.76 | 0.65 | 0.7 | 264 |
104
- | 7 | 0.81 | 0.79 | 0.8 | 189 |
105
- | 8 | 0.8 | 0.8 | 0.8 | 121 |
106
- | 9 | 0.78 | 0.83 | 0.81 | 151 |
107
- | 10 | 0.72 | 0.66 | 0.69 | 304 |
108
- | 11 | 0.67 | 0.67 | 0.67 | 452 |
109
- | 12 | 0.68 | 0.72 | 0.7 | 184 |
110
- | 13 | 0.54 | 0.64 | 0.59 | 269 |
111
- | 14 | 0.78 | 0.7 | 0.74 | 212 |
112
- | 15 | 0.63 | 0.77 | 0.69 | 154 |
113
- | 16 | 0.53 | 0.57 | 0.55 | 61 |
114
- | 17 | 0.75 | 0.71 | 0.73 | 471 |
115
- | 18 | 0.62 | 0.62 | 0.62 | 301 |
116
- | 19 | 0.27 | 0.5 | 0.35 | 8 |
117
- | 20 | 0.79 | 0.82 | 0.81 | 78 |
118
- | 21 | 0 | 0 | 0 | 0 |
119
- | macro avg | 0.66 | 0.68 | 0.67 | 5227 |
120
- | weighted avg | 0.71 | 0.7 | 0.7 | 5227 |
121
 
122
  ## Inference platform
123
  This model is used by the [CAP Babel Machine](https://babel.poltextlab.com), an open-source and free natural language processing tool, designed to simplify and speed up projects for comparative research.
 
90
  We also incorporated an EarlyStoppingCallback in the process with a patience of 2 epochs.
91
 
92
  ## Model performance
93
+ The model was evaluated on a test set of 6309 examples (10% of the available data).<br>
94
+ Model accuracy is **0.69**.
95
  | label | precision | recall | f1-score | support |
96
  |:-------------|------------:|---------:|-----------:|----------:|
97
+ | 0 | 0.65 | 0.6 | 0.62 | 621 |
98
+ | 1 | 0.71 | 0.68 | 0.69 | 473 |
99
+ | 2 | 0.79 | 0.73 | 0.76 | 247 |
100
+ | 3 | 0.77 | 0.71 | 0.74 | 156 |
101
+ | 4 | 0.68 | 0.58 | 0.63 | 383 |
102
+ | 5 | 0.79 | 0.82 | 0.8 | 351 |
103
+ | 6 | 0.71 | 0.78 | 0.74 | 329 |
104
+ | 7 | 0.81 | 0.79 | 0.8 | 216 |
105
+ | 8 | 0.78 | 0.75 | 0.76 | 157 |
106
+ | 9 | 0.87 | 0.78 | 0.83 | 272 |
107
+ | 10 | 0.61 | 0.68 | 0.64 | 315 |
108
+ | 11 | 0.61 | 0.74 | 0.67 | 487 |
109
+ | 12 | 0.72 | 0.7 | 0.71 | 145 |
110
+ | 13 | 0.69 | 0.6 | 0.64 | 346 |
111
+ | 14 | 0.75 | 0.69 | 0.72 | 359 |
112
+ | 15 | 0.69 | 0.65 | 0.67 | 189 |
113
+ | 16 | 0.36 | 0.47 | 0.41 | 55 |
114
+ | 17 | 0.68 | 0.73 | 0.71 | 618 |
115
+ | 18 | 0.61 | 0.68 | 0.64 | 469 |
116
+ | 19 | 0 | 0 | 0 | 18 |
117
+ | 20 | 0.73 | 0.75 | 0.74 | 102 |
118
+ | 21 | 0 | 0 | 0 | 1 |
119
+ | macro avg | 0.64 | 0.63 | 0.63 | 6309 |
120
+ | weighted avg | 0.7 | 0.69 | 0.69 | 6309 |
121
 
122
  ## Inference platform
123
  This model is used by the [CAP Babel Machine](https://babel.poltextlab.com), an open-source and free natural language processing tool, designed to simplify and speed up projects for comparative research.