--- library_name: setfit tags: - setfit - sentence-transformers - text-classification - generated_from_setfit_trainer metrics: - accuracy widget: - text: >- C/ BORGES BLANQUES - BORDETA : Com a veina del C/ BORGES BLANQUES de Lleida estic indignada per la falta de neteja que tenim en aquesta zona de la Bordeta. El camps segons tenim entès propietat de l'ajuntament de Lleida estan amb herbes seques de 2m d'alt que com ja va passar aquest cap de setmana es van cremar on hi ha zones habitades. La sequia que passa per alli plena d'herbes que tampoc es veu on comença o on acaba. I esperar que no plogui i se'ns inundi per la falta de neteja els parkings de la zona com ja ha passat en alguna ocasió que la sèquia no pot soportar tanta aigua i bruticia. Agrairem es neteji aquesta zona abans de tenir que lamentar mals majors. - text: >- HAN TRET LES RAJOLES DEL CARRER DE DAVANT DE LA COMUNITAT : Estan fent obres a la comunitat de davant (bloc nº 22) i els operaris de la llum (es desconeix l'empresa), com el transformador es troba soterrat just davant de la comunitat Princep de Viana 15, han tingut que obrir la vorera. Han tret les rajoles (van sortir senceres) i no les han tornat a col·locar. Es un perill, perquè la gent trepitja al caminar, o inclús al sortir de la nostra comunitat. Sol·licito sis plau, si les poden tornar a col·locar, ja que a part de que no queda estètic, es un perill i més per la gent gran. adjunto fotografies. Gràcies. - text: >- Mateicula estiu de petits : Bona tarda, Un cop adjudicada la plaça a estiu de petits fins quan es por formalitzar la matricula? Gràcies - text: >- Fer peatonal la via central de doctor fleming : Bon dia fa temps que pensem que seria bona idea fer un petit passeig a l'Avinguda Dr Fleming, la part per sobre Passeig de Ronda. Actualment hi ha 4 carrils (2 centrals i 2 laterals) inicialment imagino que pel Camp d'Esports (tot i que no hi ha massa aforament) si els centrals fossin peatonals s'aconseguiria conectar la part de Ciutat Jardí i zona Ricard Vinyes amb un passeig agradable. Al carrer hi ha 3/4 locals buits i potser es es motivaria a fer negocis (tipus cafeteria o comerç de proximitat). A Lleida ens falten zones agradables on caminar i hi hauria una bona oportunitat en aquesta avinguda. Moltes gràcies per llegir-nos!!! Helena i Òscar - text: >- Ticket guardería escola bressol Municipal Valiet : Bon dia, Necessitaria saber si la gestio de l’escola bresssol Vailet la porteu centralitzadament. La meva empresa fa un pagament mensual de 73€ a l’escola en nom de la meva filla que dita escola afirma no haver rebut. Gràcies pipeline_tag: text-classification inference: true base_model: ibaucells/RoBERTa-ca-CaWikiTC model-index: - name: SetFit with ibaucells/RoBERTa-ca-CaWikiTC results: - task: type: text-classification name: Text Classification dataset: name: Unknown type: unknown split: test metrics: - type: accuracy value: 0.4235294117647059 name: Accuracy --- # SetFit with ibaucells/RoBERTa-ca-CaWikiTC This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [ibaucells/RoBERTa-ca-CaWikiTC](https://huggingface.co/ibaucells/RoBERTa-ca-CaWikiTC) as the Sentence Transformer embedding model. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) 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:** [ibaucells/RoBERTa-ca-CaWikiTC](https://huggingface.co/ibaucells/RoBERTa-ca-CaWikiTC) - **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:** 17 classes ### Model Sources - **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit) - **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055) - **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit) ### Model Labels | Label | Examples | |:------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | 0 | | | 1 | | | 2 | | | 3 | | | 4 | | | 5 | | | 6 | | | 7 | | | 8 | | | 9 | | | 10 | | | 11 | | | 12 | | | 13 | | | 14 | | | 15 | | | 16 | | ## Evaluation ### Metrics | Label | Accuracy | |:--------|:---------| | **all** | 0.4235 | ## 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("adriansanz/test8") # Run inference preds = model("Mateicula estiu de petits : Bona tarda, Un cop adjudicada la plaça a estiu de petits fins quan es por formalitzar la matricula? Gràcies") ``` ## Training Details ### Training Set Metrics | Training set | Min | Median | Max | |:-------------|:----|:-------|:----| | Word count | 5 | 5.2353 | 7 | | Label | Training Sample Count | |:------|:----------------------| | 0 | 8 | | 1 | 8 | | 2 | 8 | | 3 | 8 | | 4 | 8 | | 5 | 8 | | 6 | 8 | | 7 | 8 | | 8 | 8 | | 9 | 8 | | 10 | 8 | | 11 | 8 | | 12 | 8 | | 13 | 8 | | 14 | 8 | | 15 | 8 | | 16 | 8 | ### Training Hyperparameters - batch_size: (8, 8) - num_epochs: (1, 1) - max_steps: -1 - sampling_strategy: oversampling - 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 - seed: 42 - eval_max_steps: -1 - load_best_model_at_end: False ### Training Results | Epoch | Step | Training Loss | Validation Loss | |:-------:|:--------:|:-------------:|:---------------:| | 0.0005 | 1 | 0.4258 | - | | 0.0230 | 50 | 0.201 | - | | 0.0460 | 100 | 0.194 | - | | 0.0689 | 150 | 0.237 | - | | 0.0919 | 200 | 0.1165 | - | | 0.1149 | 250 | 0.0622 | - | | 0.1379 | 300 | 0.0904 | - | | 0.1608 | 350 | 0.0045 | - | | 0.1838 | 400 | 0.0188 | - | | 0.2068 | 450 | 0.0025 | - | | 0.2298 | 500 | 0.0017 | - | | 0.2528 | 550 | 0.0014 | - | | 0.2757 | 600 | 0.0013 | - | | 0.2987 | 650 | 0.0014 | - | | 0.3217 | 700 | 0.0027 | - | | 0.3447 | 750 | 0.0014 | - | | 0.3676 | 800 | 0.0007 | - | | 0.3906 | 850 | 0.0014 | - | | 0.4136 | 900 | 0.0011 | - | | 0.4366 | 950 | 0.0011 | - | | 0.4596 | 1000 | 0.0017 | - | | 0.4825 | 1050 | 0.0007 | - | | 0.5055 | 1100 | 0.001 | - | | 0.5285 | 1150 | 0.0008 | - | | 0.5515 | 1200 | 0.0005 | - | | 0.5744 | 1250 | 0.0009 | - | | 0.5974 | 1300 | 0.0008 | - | | 0.6204 | 1350 | 0.0013 | - | | 0.6434 | 1400 | 0.0008 | - | | 0.6664 | 1450 | 0.001 | - | | 0.6893 | 1500 | 0.0006 | - | | 0.7123 | 1550 | 0.0008 | - | | 0.7353 | 1600 | 0.0006 | - | | 0.7583 | 1650 | 0.0005 | - | | 0.7812 | 1700 | 0.0006 | - | | 0.8042 | 1750 | 0.0006 | - | | 0.8272 | 1800 | 0.001 | - | | 0.8502 | 1850 | 0.0005 | - | | 0.8732 | 1900 | 0.0007 | - | | 0.8961 | 1950 | 0.0009 | - | | 0.9191 | 2000 | 0.0005 | - | | 0.9421 | 2050 | 0.0005 | - | | 0.9651 | 2100 | 0.0005 | - | | 0.9881 | 2150 | 0.0005 | - | | **1.0** | **2176** | **-** | **0.2021** | | 0.0005 | 1 | 0.0004 | - | | 0.0230 | 50 | 0.0006 | - | | 0.0460 | 100 | 0.0005 | - | | 0.0689 | 150 | 0.0005 | - | | 0.0919 | 200 | 0.0004 | - | | 0.1149 | 250 | 0.0005 | - | | 0.1379 | 300 | 0.0004 | - | | 0.1608 | 350 | 0.0018 | - | | 0.1838 | 400 | 0.0005 | - | | 0.2068 | 450 | 0.0003 | - | | 0.2298 | 500 | 0.0003 | - | | 0.2528 | 550 | 0.0003 | - | | 0.2757 | 600 | 0.0002 | - | | 0.2987 | 650 | 0.0003 | - | | 0.3217 | 700 | 0.0004 | - | | 0.3447 | 750 | 0.0002 | - | | 0.3676 | 800 | 0.0002 | - | | 0.3906 | 850 | 0.0006 | - | | 0.4136 | 900 | 0.0002 | - | | 0.4366 | 950 | 0.0002 | - | | 0.4596 | 1000 | 0.0004 | - | | 0.4825 | 1050 | 0.0004 | - | | 0.5055 | 1100 | 0.0003 | - | | 0.5285 | 1150 | 0.0002 | - | | 0.5515 | 1200 | 0.0002 | - | | 0.5744 | 1250 | 0.0002 | - | | 0.5974 | 1300 | 0.0002 | - | | 0.6204 | 1350 | 0.0003 | - | | 0.6434 | 1400 | 0.0001 | - | | 0.6664 | 1450 | 0.0003 | - | | 0.6893 | 1500 | 0.0002 | - | | 0.7123 | 1550 | 0.0003 | - | | 0.7353 | 1600 | 0.0002 | - | | 0.7583 | 1650 | 0.0002 | - | | 0.7812 | 1700 | 0.0002 | - | | 0.8042 | 1750 | 0.0002 | - | | 0.8272 | 1800 | 0.0002 | - | | 0.8502 | 1850 | 0.0002 | - | | 0.8732 | 1900 | 0.0003 | - | | 0.8961 | 1950 | 0.0003 | - | | 0.9191 | 2000 | 0.0003 | - | | 0.9421 | 2050 | 0.0002 | - | | 0.9651 | 2100 | 0.0002 | - | | 0.9881 | 2150 | 0.0002 | - | | **1.0** | **2176** | **-** | **0.1685** | * The bold row denotes the saved checkpoint. ### Framework Versions - Python: 3.10.12 - SetFit: 1.0.3 - Sentence Transformers: 2.6.1 - Transformers: 4.38.2 - PyTorch: 2.2.1+cu121 - Datasets: 2.18.0 - Tokenizers: 0.15.2 ## Citation ### BibTeX ```bibtex @article{https://doi.org/10.48550/arxiv.2209.11055, doi = {10.48550/ARXIV.2209.11055}, url = {https://arxiv.org/abs/2209.11055}, author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren}, keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences}, title = {Efficient Few-Shot Learning Without Prompts}, publisher = {arXiv}, year = {2022}, copyright = {Creative Commons Attribution 4.0 International} } ```