|
|
--- |
|
|
tags: |
|
|
- setfit |
|
|
- sentence-transformers |
|
|
- text-classification |
|
|
- generated_from_setfit_trainer |
|
|
widget: |
|
|
- text: Etunimi Etunimi menetkö noin vaan takuuseen, ettei sodan johdosta näin käy? |
|
|
Ite en kyllä menis 100% sanomaan mitään mihin liittyy Putin ja Putinin sota |
|
|
- text: Kohta on lisää lapsia sairaalassa koronan vuoksi ☹ |
|
|
- text: Etunimi Sukunimi pyöräily sekä kävely ovat hyvää liikuntaa |
|
|
- text: Etunimi Sukunimi Niin.. Nuo todelliset tartunyamäärät voivat olla ihan mitä |
|
|
tahansa. Mihinkään rajoitustoimiin ei tarvitsisi ryhtyä. Ihmiset voivat itse pitää |
|
|
huolta itsestää, ja valtion tehtävä on pitää huolta siitä että hoitokapasiteetti |
|
|
riittää. Tällä hetkellä meillä ei ole mitään hätää. Koko Suomessa tehohoidossa |
|
|
koronan vuoksi on noin 2p ihmistä. Tehohoitopaikkoja siis riittää vielä vaikka |
|
|
ja kuinka jos tarvetta. Korostan, että edelleenkin ovat turvavälit, hyvä hygienia |
|
|
ja turhien kontaktien välttäminen kaikkein tärkeintä. Mitään ei tarvitsisi rajoittaa, |
|
|
jollei ihmiset olisi niin helvetin tyhmiä, että osaisivat ajatella ihan omilla |
|
|
aivoillaan, eikä valtion tarvitsisi heitä opastaa kädestä pitäen kuten jotain |
|
|
pieniä lapsia. |
|
|
- text: Etunimi hallituksella pitää kuitenkin olla jokin pohja johon perustavat päätöksensä. |
|
|
Poikkeustilaa ei voi loputtomiin jatkaa vain mutulla, jolloin heidän on kuunneltava |
|
|
aiheen ammattilaisia. |
|
|
metrics: |
|
|
- metric |
|
|
pipeline_tag: text-classification |
|
|
library_name: setfit |
|
|
inference: true |
|
|
base_model: TurkuNLP/bert-base-finnish-cased-v1 |
|
|
model-index: |
|
|
- name: SetFit with TurkuNLP/bert-base-finnish-cased-v1 |
|
|
results: |
|
|
- task: |
|
|
type: text-classification |
|
|
name: Text Classification |
|
|
dataset: |
|
|
name: Unknown |
|
|
type: unknown |
|
|
split: test |
|
|
metrics: |
|
|
- type: metric |
|
|
value: 0.8718523964493842 |
|
|
name: Metric |
|
|
--- |
|
|
|
|
|
# Detect Actions in Asynchronous Conversation Comments |
|
|
|
|
|
## SetFit with TurkuNLP/bert-base-finnish-cased-v1 |
|
|
|
|
|
This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification of actions in asynchronous conversation. This particular model detects if a comment includes a statement or not. The configuration of the model is that the model is based on averaged annotations (from 3 annotators). Metric evaluations are based on conservative ground truth (see paper). This SetFit model uses TurkuNLP/bert-base-finnish-cased-v1 as the Sentence Transformer embedding model (using word embeddings). A LogisticRegression instance is used for classification. |
|
|
|
|
|
The model has been trained using an efficient few-shot learning technique that involves: |
|
|
|
|
|
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning. |
|
|
2. Training a classification head with features from the fine-tuned Sentence Transformer. |
|
|
|
|
|
## Model Details |
|
|
|
|
|
### Model Description |
|
|
- **Model Type:** SetFit |
|
|
- **Sentence Transformer body:** [TurkuNLP/bert-base-finnish-cased-v1](https://huggingface.co/TurkuNLP/bert-base-finnish-cased-v1) |
|
|
- **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance |
|
|
- **Maximum Sequence Length:** 512 tokens |
|
|
- **Number of Classes:** 2 classes (action present yes/no) |
|
|
<!-- - **Training Dataset:** [Unknown](https://huggingface.co/datasets/unknown) --> |
|
|
<!-- - **Language:** Unknown --> |
|
|
<!-- - **License:** Unknown --> |
|
|
|
|
|
### Model Sources |
|
|
|
|
|
- **Repository:** [GitHub](https://github.com/henniina/Detecting-paired-actions) |
|
|
- **Paper:** Paakki, H., Toivanen, P. and Kajava K. (2025). Implicit and Indirect: Detecting Face-threatening and Paired Actions in Asynchronous Online Conversations. Northern European Journal of Language Technology (NEJLT), 11(1), pp. 58-83. |
|
|
- |
|
|
### Model Labels |
|
|
| Label | Examples | |
|
|
|:------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| |
|
|
| 0 | <ul><li>'Etunimi Sukunimi miten luulet tilanteen parantuneen kun sairaala- ja tehohoito potilaiden määrä on vain kasvanut silloisesta?\nOlet niin totaalisen puusilmäinen ja hallirusvihan vallassa, että tätä on turha jatkaa pitemmälle. Pysy terveenä ja rauhallista joulua!'</li><li>'Etunimi Sukunimi perustuslakia ei ole rikottu niissä asioissa mitä convoypellet väitti, kaikki mitä kaverit väittää ei ole totta .'</li><li>'Takaisin hiilivoimaan?'</li></ul> | |
|
|
| 1 | <ul><li>'"Hylkiö" unionin toimesta johon ei kuulu.'</li><li>'Etunimi Sukunimi en nyt varsinaisesti pelkästään tuota aihetta tarkoittanutkaan. Sekin on kuitenkin vähintään kyseenalaista, koska kyseessä ei ole valmis tuote, vaan hätämyyntiluvalla käytössä oleva ruiske, ja sen seurauksena on niinikään perusoikeudellinen terveydenhuollon taso turvaamattomalla tasolla.'</li><li>'Mikään ei ole niin varmaa kuin epävarma. KUKAAN ei millään voi tietää mitä tapahtuu koronan tai ylipäätään minkään suhteen. Joka muuta väittää on typerys...'</li></ul> | |
|
|
|
|
|
## Evaluation |
|
|
|
|
|
### Metrics |
|
|
| Label | Metric | |
|
|
|:--------|:-------| |
|
|
| **5-fold cross-validated F1** | 0.74 | |
|
|
|
|
|
## Uses |
|
|
|
|
|
### Direct Use for Inference |
|
|
|
|
|
First install the SetFit library: |
|
|
|
|
|
```bash |
|
|
pip install setfit |
|
|
``` |
|
|
|
|
|
Then you can load this model and run inference. |
|
|
|
|
|
```python |
|
|
from setfit import SetFitModel |
|
|
|
|
|
# Download from the 🤗 Hub |
|
|
model = SetFitModel.from_pretrained("Finnish-actions/SetFit-FinBERT1-Avg-statement") |
|
|
# Run inference |
|
|
preds = model("Kohta on lisää lapsia sairaalassa koronan vuoksi ☹") |
|
|
``` |
|
|
|
|
|
|
|
|
### Downstream Use |
|
|
|
|
|
NB. This model has been trained on data coming from Finnish language asynchronous conversations under crisis related news on Facebook. This specific model has been trained to detect whether a comment includes a question or not. It reflects only one of our annotators' label interpretations, so the best use of our models (see our paper) would be to combine a set of models we provide on our Huggingface (Finnish-actions), and use a model ensemble to provide label predictions. It needs to be noted also that the model may not be well applicable outside of its empirical context, so in downstream applications, one should always conduct an evaluation of the model applicability using manually annotated data from that specific context (see our paper for annotation instructions). |
|
|
|
|
|
## Out-of-Scope Use |
|
|
|
|
|
Please use this model only for action detection and analysis. Uses of this model and the involved data for generative purposes (e.g. NLG) is prohibited. |
|
|
|
|
|
|
|
|
### Bias, Risks and Limitations |
|
|
|
|
|
Note that the model may produce errors. Due to the size of the training dataset, model may not generalize very well even for other novel topics within the same context. Note that model predictions should not be regarded as final judgments e.g. for online moderation purposes, but each case should also be regarded individually if using model predictions to support moderation. Also, the annotations only reflect three (though experienced) annotators' interpretations, so there might be perspectives on data intepretation that have not been taken into account here. |
|
|
|
|
|
If model is used to support moderation on social media, we recommend that final judgments should always be left for human moderators. |
|
|
|
|
|
|
|
|
<!-- |
|
|
### Out-of-Scope Use |
|
|
|
|
|
*List how the model may foreseeably be misused and address what users ought not to do with the model.* |
|
|
--> |
|
|
|
|
|
<!-- |
|
|
## Bias, Risks and Limitations |
|
|
|
|
|
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.* |
|
|
--> |
|
|
|
|
|
<!-- |
|
|
### Recommendations |
|
|
|
|
|
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.* |
|
|
--> |
|
|
|
|
|
## Training Details |
|
|
|
|
|
### Training Set Metrics |
|
|
| Training set | Min | Median | Max | |
|
|
|:-------------|:----|:--------|:----| |
|
|
| Word count | 1 | 19.9323 | 213 | |
|
|
|
|
|
| Label | Training Sample Count | |
|
|
|:------|:----------------------| |
|
|
| 0 | 218 | |
|
|
| 1 | 624 | |
|
|
|
|
|
<!-- |
|
|
### Training Hyperparameters |
|
|
- batch_size: (16, 16) |
|
|
- num_epochs: (4, 4) |
|
|
- max_steps: -1 |
|
|
- sampling_strategy: oversampling |
|
|
- num_iterations: 6 |
|
|
- body_learning_rate: (2e-05, 1e-05) |
|
|
- head_learning_rate: 0.01 |
|
|
- loss: CosineSimilarityLoss |
|
|
- distance_metric: cosine_distance |
|
|
- margin: 0.25 |
|
|
- end_to_end: False |
|
|
- use_amp: False |
|
|
- warmup_proportion: 0.1 |
|
|
- l2_weight: 0.01 |
|
|
- seed: 42 |
|
|
- evaluation_strategy: epoch |
|
|
- eval_max_steps: -1 |
|
|
- load_best_model_at_end: False |
|
|
|
|
|
### Training Results |
|
|
| Epoch | Step | Training Loss | Validation Loss | |
|
|
|:------:|:----:|:-------------:|:---------------:| |
|
|
| 0.0016 | 1 | 0.2335 | - | |
|
|
| 0.0791 | 50 | 0.2607 | - | |
|
|
| 0.1582 | 100 | 0.238 | - | |
|
|
| 0.2373 | 150 | 0.1995 | - | |
|
|
| 0.3165 | 200 | 0.1685 | - | |
|
|
| 0.3956 | 250 | 0.0844 | - | |
|
|
| 0.4747 | 300 | 0.0443 | - | |
|
|
| 0.5538 | 350 | 0.0202 | - | |
|
|
| 0.6329 | 400 | 0.0151 | - | |
|
|
| 0.7120 | 450 | 0.0103 | - | |
|
|
| 0.7911 | 500 | 0.0089 | - | |
|
|
| 0.8703 | 550 | 0.0098 | - | |
|
|
| 0.9494 | 600 | 0.0043 | - | |
|
|
| 1.0 | 632 | - | 0.1711 | |
|
|
| 1.0285 | 650 | 0.0007 | - | |
|
|
| 1.1076 | 700 | 0.0002 | - | |
|
|
| 1.1867 | 750 | 0.0006 | - | |
|
|
| 1.2658 | 800 | 0.0001 | - | |
|
|
| 1.3449 | 850 | 0.0012 | - | |
|
|
| 1.4241 | 900 | 0.0006 | - | |
|
|
| 1.5032 | 950 | 0.0001 | - | |
|
|
| 1.5823 | 1000 | 0.0001 | - | |
|
|
| 1.6614 | 1050 | 0.0002 | - | |
|
|
| 1.7405 | 1100 | 0.0001 | - | |
|
|
| 1.8196 | 1150 | 0.0001 | - | |
|
|
| 1.8987 | 1200 | 0.0001 | - | |
|
|
| 1.9778 | 1250 | 0.0001 | - | |
|
|
| 2.0 | 1264 | - | 0.1870 | |
|
|
| 2.0570 | 1300 | 0.0001 | - | |
|
|
| 2.1361 | 1350 | 0.0001 | - | |
|
|
| 2.2152 | 1400 | 0.0001 | - | |
|
|
| 2.2943 | 1450 | 0.0001 | - | |
|
|
| 2.3734 | 1500 | 0.0001 | - | |
|
|
| 2.4525 | 1550 | 0.0 | - | |
|
|
| 2.5316 | 1600 | 0.0001 | - | |
|
|
| 2.6108 | 1650 | 0.0001 | - | |
|
|
| 2.6899 | 1700 | 0.0001 | - | |
|
|
| 2.7690 | 1750 | 0.0001 | - | |
|
|
| 2.8481 | 1800 | 0.0 | - | |
|
|
| 2.9272 | 1850 | 0.0001 | - | |
|
|
| 3.0 | 1896 | - | 0.1924 | |
|
|
| 3.0063 | 1900 | 0.0 | - | |
|
|
| 3.0854 | 1950 | 0.0 | - | |
|
|
| 3.1646 | 2000 | 0.0001 | - | |
|
|
| 3.2437 | 2050 | 0.0 | - | |
|
|
| 3.3228 | 2100 | 0.0 | - | |
|
|
| 3.4019 | 2150 | 0.0 | - | |
|
|
| 3.4810 | 2200 | 0.0 | - | |
|
|
| 3.5601 | 2250 | 0.0001 | - | |
|
|
| 3.6392 | 2300 | 0.0001 | - | |
|
|
| 3.7184 | 2350 | 0.0 | - | |
|
|
| 3.7975 | 2400 | 0.0 | - | |
|
|
| 3.8766 | 2450 | 0.0 | - | |
|
|
| 3.9557 | 2500 | 0.0 | - | |
|
|
| 4.0 | 2528 | - | 0.1996 | |
|
|
--> |
|
|
|
|
|
### Framework Versions |
|
|
- Python: 3.11.9 |
|
|
- SetFit: 1.1.3 |
|
|
- Sentence Transformers: 3.2.0 |
|
|
- Transformers: 4.44.0 |
|
|
- PyTorch: 2.4.0+cu124 |
|
|
- Datasets: 2.21.0 |
|
|
- Tokenizers: 0.19.1 |
|
|
|
|
|
## Citation |
|
|
|
|
|
### BibTeX |
|
|
```bibtex |
|
|
@article{paakki-implicit-indirect, |
|
|
doi = {https://doi.org/10.3384/nejlt.2000-1533.2025.5980}, |
|
|
url = {https://nejlt.ep.liu.se/article/view/5980}, |
|
|
author = {Paakki, Henna and Toivanen, Pihla and Kajava, Kaisla}, |
|
|
title = {Implicit and Indirect: Detecting Face-threatening and Paired Actions in Asynchronous Online Conversations}, |
|
|
publisher = {Northern European Journal of Language Technology (NEJLT)}, |
|
|
volume= {11}, |
|
|
number= {1}, |
|
|
year = {2025} |
|
|
} |
|
|
|
|
|
``` |
|
|
|
|
|
<!-- |
|
|
## Glossary |
|
|
|
|
|
*Clearly define terms in order to be accessible across audiences.* |
|
|
--> |
|
|
|
|
|
<!-- |
|
|
## Model Card Authors |
|
|
|
|
|
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.* |
|
|
--> |
|
|
|
|
|
<!-- |
|
|
## Model Card Contact |
|
|
|
|
|
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.* |
|
|
--> |
